Home

A Computational Platform for Assessing the Impact of Alternative

Contents

1. 46 Web Service User Manual e WebService e WebService Aspnet e WebService Common e WebService Common Logic e WebService Common WS e WebService Self e WebService WorkerBot e WebService sin Each project folder contains a csproj project file for compilation The root of the folder has a s1n file that compiles the whole application WebService sln To compile the code you need to have a version of Visual Studio or MonoDevelop that supports C 4 0 projects and NuGet packages these should be automatically restored when building Visual Studio 2010 or newer and the latest version of MonoDevelop Mono gt 3 2 8 should be able to compile the project The solution uses MSBuild for compilation It is a build automation tool from Microsoft that uses an XML based project file format Besides being able to compile from Visual Studio or MonoDevelop you can also compile using the command line using MSBuild in Windows or XBuild in any platform supported by Mono To compile the project simply run msbuild exe WebService sln xbuild ex WebService sln This command compiles the solution in its default configuration which is Debug To compile in Release mode use msbuild exe WebService sln property Configuration Release xbuild ex WebService sln property Configuration Release Note Compiled binaries with either MSBuild or XBuild should work without modif
2. Figure 3 7 Markdown example 48 Markdown http en wikipedia org wiki Markdown 32 A Website for Alternative Splicing Analysis File management is done using the files browser where a user can browse public files and project s specific files as seen in Figure 3 8 He can also delete files and add descriptions to them to better know what they are and where they came from Size Creation Date Modify Date 14 14 09 37 Description Rename bamfile bam 15 06 11 14 13 04 41 Description Rename Delete atffile gtf 3 5K 15 014 11 04 014 15 50 21 Description Rename Delete smallbam bam 893 075 KB 15 06 2 14 15 50 21 Description Rename Delete Figure 3 8 File browser The web service and the worker bot can be configured using App config files located in the program s folders These can be used to configure server name port and the location of the data folders Before running the web service and the worker bot these must be configured with valid paths to a folder to store the data files and a folder for the stdio files Default folder names are used in case these are not set The web service is written in C using the Nancy web framework a lightweight web framework designed to be simple modular and easy to extend where a user can choose which functionalities to use Authentication is provided by Nancy s forms authentication module The view engine used to process the HTML pages is the Razor view engine
3. In today s world many of the interactions with the internet are made from touch screens whether phones or tablets Therefore it is necessary to have the solution work well with touch centric devices by avoiding certain user interface choices such as small buttons and relying on the mouse to perform specific tasks such as mouse hover tooltips e The solution must have the notion of users and authentication It should not be possible to access any of the web service s functionality without authenticating first with the system New user accounts should also be validated by an administrator account before being allowed to access the system e The solution must allow the creation of projects A project is a mere description of an experiment where a user can try new things and have its own personal areas to work in e The solution must allow the creation of jobs A job is a background task that is to be executed by the solution and have its output saved for later use e The solution must be able to run jobs in a different computer In order to have a job run in a more powerful computer or have more than one computer available to run jobs the solution should be able to send jobs to another computer and have the results sent back when done e The solution must be able to download files A user should be able to download files produced by the solution e The solution must allow file descriptions A user should
4. Simple file viewer to view and download files stored in the server e Jobs Manages jobs to be run by worker bots e Projects Manages projects describing experiments e Users User account management 54 Web Service User Manual A 3 4 Clusters Cluster accounts are used by worker bots to access the web service You can list cluster accounts and check their status by viewing the last activity Cluster accounts when in use poll the web service every 10 seconds which means that if a cluster account is idle it should have a last activity below 10 seconds Figure A 7 Cluster Accounts Login Creation Date Last Activity Cluster 22 06 2014 13 20 08 5 sec S Delete Figure A 7 Cluster accounts listing A 3 4 1 Create Cluster You can create a cluster account from the clusters menu You need to enter the following fields to create an account Figure A 8 e Login Unique account login Must be at least 3 characters long e Password Password used to authenticate into the web service Must be at least 4 characters long Clusters Create Cluster Account Add Cancel Figure A 8 Add cluster account 55 Web Service User Manual A 3 5 Files You can use the files menu to browse files in the server Figure A 9 Size Creation Date Modify Date 59 folder 15 014 1 5 1 7 Description Rename Delete bam ile bam 10 5B 15 06 3 03 01 01 2014 13 Description Rename gtrfile gtf 370 5 KB 15 06 2014
5. List Users Add User Pending Users 4 Figure A 19 Pending users menu You can clear the pending state by clicking the clear pending button in the users listing Figure A 20 Users Login Name Creation Date admin Administrator 31 12 2013 0 00 Details Useri John 22 06 2014 15 54 01 Clear Pending Details Figure A 20 Users listing 64
6. Name of the folder to store the data files Default is data StdioFolder Name of the folder to store the command s stdio files Default is stdio These settings can be configured in the WebService WorkerBot exe config file Figure A 2 lt xml version 1 0 encoding utf 8 gt lt configuration gt lt appSettings gt lt add key URL value http localhost 8080 gt lt add key Login value Cluster1 gt lt add key PassHash value 5863d9e4cbdf522eaa62e0747fceb1c5b249ba13 gt lt add key DataFolder value data gt lt add key StdioFolder value stdio gt lt add key LocalBot value true gt lt appSettings gt lt configuration gt Figure A 2 Worker bot example configuration file If the worker bot is running in the same computer it is recommended that both the web service and worker bot share the data and stdio folders To do this make sure that both configuration files point to the same folders on the disk or create symbolic links in one of the program s root folder that link to the other program s folders You also need to set the LocalBot parameter to true Both folders should have read and write permission from the operating system 51 Web Service User Manual A 3 User Guide You can use the web service by first opening the web service s URL using a web browser In order to access the web service you must first login into the web service or create a ne
7. TopHat discovering splice junctions with RNA Seq Bioinformatics no 25 9 1105 11 http www ncbi nlm nih gov pubmed 19289445 doi 10 1093 bioinformatics btp 120 Trapnell C B A Williams G Pertea A Mortazavi G Kwan M J van Baren S L Salzberg B J Wold and L Pachter 2010 Transcript assembly and quantification by RNA Seq reveals unannotated transcripts and isoform switching during cell differentiation Nat Biotechnol no 28 5 511 5 http www ncbi nlm nih gov pubmed 20436464 43 References http www ncbi nlm nih gov pmc articles PMC3146043 pdf nihms190938 pdf doi 10 1038 nbt 1621 Venables Julian P 2004 Aberrant and Alternative Splicing in Cancer Cancer Research no 64 21 7647 7654 http cancerres aacrjournals org content 64 21 7647 abstract doi 10 1158 0008 5472 can 04 1910 Zdobnov Evgeni M and Rolf Apweiler 2001 InterProScan an integration platform for the signature recognition methods in InterPro Bioinformatics no 17 9 847 848 http bioinformatics oxfordjournals org content 17 9 847 abstract doi 10 1093 bioinformatics 17 9 847 44 Appendix A A Web Service User Manual A 1 Introduction This document describes the necessary steps to have the web service up and running A 2 Installation A 2 1 Prerequisites The solution is written in C and targets the NET framework 4 0 In order to run the solution you need one of
8. WebService WorkerBot exe config A 2 3 Installing In order to install either the web service or the worker bot all it is needed is to copy the necessary files into their destination and to run the application the exe file To run the exe on Windows simply run it as administrator or call it from an administrator command line since a user mode command line will fail to open the necessary port On Mono you must first append mono to the command line mono WebService Self ex 48 Web Service User Manual Both projects require prior configuration to work properly Configuration is achieved by using NET framework s built in application configuration files These files contain settings specific to the application and usually have the same name of the application with config appended to the end These files should have special write permission from the operating system even if they are running in an account without writing permissions in the application s folder This behaviour is guaranteed to happen on Windows but might not be on other operating systems A 2 3 1 Web Service WebService Self ex WebService Self exe config The web service supports the h or help parameters that print the application s supported configuration parameters and their default values e URL URL to use to start the web service The HTTP protocol string must be present otherwise an exception will be thrown Default is http l
9. 11 13 38 04 06 2014 15 50 2 Description Rename smallbam bam 893 075 KB 15 06 2 04 06 2014 15 50 21 Description Rename Figure A 9 File browser You can perform the following actions e Add a description to a file Figure A 11 e Edit that description Figure A 10 e Rename a file Figure A 12 e Delete a file Figure A 13 File descriptions use Markdown and come with an editor for easier editing Edit description Figure A 10 Edit file description 56 Web Service User Manual Figure A 11 File description Enter the new file name for bamfile bam Cancel Es Figure A 12 Rename file dialog cuco Figure A 13 Delete file confirmation dialog 57 Web Service User Manual Folders are used by projects for their own files and identify themselves with the project s name for easy identification Figure A 14 You can also navigate between folders using the folder bar at the top or by using the links present in sub folders e Root folder e Up one folder Home Files 53a6e43bf525bc006046e759 Files Name Creation Date Modify Date atrfile atf 3 5 KB 5 20 OS 04 0 314 15 50 21 Description smallbambam 893 075 KB 15 06 2014 14 09 37 04 06 2014 15 50 21 Description Figure A 14 Project s file browser 58 A 3 6 Web Service User Manual Jobs Jobs are actions that are to be performed by an available worker bot There currently are these types of jobs
10. Hash Algorithm Structured Query Language Standard Error Standard Input Standard Input Output Standard Output Support Vector Machine Uniform Resource Locator Extensible Markup Language XIV Chapter 1 1 Introduction This chapter gives an introduction to the problem being addresses by the thesis work and describes the solution adopted and implemented We also provide a motivation for the work done and summarize the contributions 1 1 Context Cancer is a disease that affects millions of individuals around the world every year Although it is known by this single name it may have different causes One such cause is based on an abnormal working of the genetic mechanisms Aberrant alternative splicing has been recognised as having a very important role in cancer development Sette Ladomery and Ghigna 2013 Venables 2004 The emergence of high throughput sequencing especially with the next generation sequencing NGS techniques has brought substantial advancements in cancer genomics research Newer NGS techniques are faster cheaper and with advancements in computing allow us to analyse the process using cheap off the shelf hardware The current pace of advancements of this technology made it possible to assay with high depth of coverage the DNA and the transcriptomes RNA Seq of tumour samples and cancer cell lines Aberrant RNA processing of genes plays an important role in cancer Along with the advancements on sequencing
11. On the client side Twitter Bootstrap is used as a front end framework for creating common interface components such as forms buttons and lists This allows us to focus on the business logic Bootstrap takes advantage of the various screen resolutions to rearrange content based on screen size Bootstrap allows us to change the appearance simply by modifying the default cascading style sheet CSS file The default template is a modified version of the cyborg theme JQuery is used for common user interface tasks on the client side The worker bot uses the RestSharp REST API to communicate with the web service All dependencies are managed using NuGet the package manager used by NET The web service was tested in both Windows and Linux machines using Mono an open source implementation of NET Nancy web framework http nancyfx org 5 Razor view engine http razorengine codeplex com 3 Twitter Bootstrap http getbootstrap com 32 Cyborg theme for bootstrap by Thomas Park http bootswatch com cyborg 5 RestSharp http restsharp org NuGet package manager http www nuget org 33 A Website for Alternative Splicing Analysis 3 3 Chapter Conclusions Due to the nature of bioinformatics tools being mostly command line based the web service is designed to be a generic job scheduling application where new jobs can be added to the web service and processed by background workers Some of the s
12. a separate work area for each project Jobs that are associated to projects will output their files on that job s work area A 3 7 1 Create Project You can create projects from the clusters menu You need to enter the following fields to create a project Figure A 17 e Name Project name e Description Project description Descriptions use Markdown and come with an editor for easier editing Projects Create Project EM OX OC Add Cancel Figure A 17 Add project 62 Web Service User Manual A 3 7 2 Project Details Once the project is created you can add jobs associated with it and view the project s files The description is useful to describe the experiment and to add links to the files and special notes regarding the experiment Figure A 18 Project Details Name Creation Date Project 1 22 06 2014 14 12 11 Description Testing Jobs to run Project files Project jobs edt Delete Figure A 18 Project details 63 Web Service User Manual A 3 8 Users User accounts are used for authentication They are created using the singed in menu link as described in the beginning of this guide A 3 8 1 Pending User When a new user is created by a guest it is created in a pending state and you can t login into the system using that account until an administrator clears the pending state Administrators are notified in the menu bar when pending users exist Figure A 19 Users Da
13. alternative splicing Classification methods that discriminate between normal and aberrant alternative splicing could also help in the explanation of the phenomena However for a proper and complete use of such approaches we need more time to enrich the information available in the ENCODE project s database that we used 42 References Kim D G Pertea C Trapnell H Pimentel R Kelley and S L Salzberg 2013 TopHat2 accurate alignment of transcriptomes in the presence of insertions deletions and gene fusions Genome Biol no 14 4 R36 http www ncbi nlm nih gov pubmed 23618408 doi 10 1186 gb 2013 14 4 r36 Kim D and S L Salzberg 2011 TopHat Fusion an algorithm for discovery of novel fusion transcripts Genome Biol no 12 8 R72 http www ncbi nlm nih gov pubmed 2 1835007 doi 10 1186 gb 2011 12 8 172 Langmead B and S L Salzberg 2012 Fast gapped read alignment with Bowtie 2 Nat Methods no 9 4 357 9 http www ncbi nlm nih gov pubmed 22388286 doi 10 1038 nmeth 1923 Langmead B M C Schatz J Lin M Pop and S L Salzberg 2009 Searching for SNPs with cloud computing Genome Biol no 10 11 R134 http www ncbi nlm nih gov pubmed 19930550 doi 10 1186 gb 2009 10 11 r134 Langmead B C Trapnell M Pop and S L Salzberg 2009 Ultrafast and memory efficient alignment of short DNA sequences to the human genome Genome Biol no 10 3 R25 http www ncbi nlm nih gov pubmed 1
14. data miners for developing statistical software and data analysis The Bioconductor project provides R packages for the analysis of genomic data such as Affymetrix and cDNA microarray object oriented data handling and analysis tools and has started to provide tools for analysis of data from next generation high throughput sequencing methods 2 4 Algorithm Evaluation Methods and Measures In order to assess the quality of the algorithms some techniques commonly used in algorithm evaluation will be applied to avoid overfitting where the testing data and the actual data vary enough to hinder the algorithms used One of these techniques consists on having two sets of data to test against one for training and another for verifying train test and cross validation Hold out is one of such techniques where a part of the data is used for training and the remaining part for testing Another technique which can help is resampling where many permutations of data are used to train the algorithms in order to add randomness to the training set by omitting data from training using a series of techniques and using the omitted data to validate the algorithm and then re training with the same data but switching the testing data with a part of the training data until all data is used in both training and testing An associated error rate will be used to assure the result in order to cope with the inherent errors that such an enormous amount o
15. description File description Rename file dialog Delete file confirmation dialog Project s file browser Add job example Job details Add project Project details Pending users menu Users listing 55 56 56 57 57 57 58 60 61 62 63 64 64 List of Tables Table 2 1 REST operations for user collection Table 2 2 REST operations for user item xl 20 20 Abbreviations API ASP BAM BED cDNA CRUD CSS DB DNA FEUP FTP GNU GPL GTF HTML HTTP IIS JS JSON mRNA MVC NGS NoSQL ORF OS OSI REST RGB RNA Application Programming Interface Active Server Pages Binary Sequence Alignment Annotation Track Format Complementary DNA Create Read Update and Delete Cascading Style Sheets Database Deoxyribonucleic Acid Faculty of Engineering of the University of Porto Faculdade de Engenharia da Universidade do Porto File Transfer Protocol GNU s not UNIX GNU General Public License Gene Transfer Format HyperText Markup Language Hypertext Transfer Protocol Internet Information Services JavaScript JavaScript Object Notation Messenger RNA Model View Controller Next Generation Sequencing Not Only SQL Open Reading Frame Operating System Open Source Initiative Representational State Transfer RGB colour model red green blue Ribonucleic Acid xiii RNA Seq SAM SHA SQL Stderr Stdin Stdio Stdout SVM URL XML RNA Sequencing Sequence Alignment Map Secure
16. of biological problems Our main motivation is to contribute to the availability of software that may speed up and facilitate the work of biologist and physicians who are involved in the fight against cancer We also take advantage of real data available on the web More specifically we are particularly motivated by the crucial help that informatics can provide to the studies requiring the analysis of massive amounts of genetic data 1 3 Project In this thesis we propose and have implemented a computational platform that may improve the analysis process by expert biologists when studying the effect of aberrant alternative splicing in the development of cancer The platform is easy to use and allows the management of users computational resources and experiments It enables the execution of tasks to be distributed among several machines running different environments Users have control over their experiments and can share information among them by making their resources public There is also implemented with the topic of aberrant alternative splicing in mind the possibility to search for information on available databases on the internet in a user transparent manner The platform provides a way to fetch store and retrieve large amounts of data that are typical in human genomic based studies General contributions of this thesis include e A general purpose architecture to run jobs in several machines running different operating systems e A we
17. protein accession name or Entrez gi s e g Q5I7T1 AG10B HUMAN 129295 but a bar separated NCBI sequence identifier e g gi 129295 will also be accepted Any arbitrary user specified sequence identifier can also be used e g CLONE00073452 but you are advised to use sufficiently long unique words in such case There should be no space between the gt and the first letter of the identifier It is recommended that all lines of text be shorter than 80 characters in length 15 FASTA format http genetics bwh harvard edu pph FASTA html 12 Basic Concepts and Survey on Technology gt gi 129295 sp P01013 OVAX CHICK GENE PROTEIN OVALBUMIN RELATED OTKDLLVSSSTDLDTTLVLVNAIYFKGMWKTAFNAE EMPFHVTKOESKPVOMMCMNNSENVATLPAE KMKILELPFASGDLSMLVLLPDEVSDLERIEKTINFEKLTEWTNPNTMEKRRVKVYLPOMKIEEKYNLTS VLMALGMTDLFIPSANLTGISSAESLKISQAVHGAFMELSEDGIEMAGSTGVIEDIKHS PESEQFRADHP FLFLIKHNPTNTIVYFGRYWSP Figure 2 8 FASTA file example 2 2 2 Genomic Assemblage Software Here we present the tools that are being considered for each phase Descriptions are taken from their respective web pages Cufflinks Cufflinks Trapnell et al 2010 Roberts Trapnell et al 2011 Roberts Pimentel et al 2011 Trapnell et al 2013 assembles transcripts estimates their abundances and tests for differential expression and
18. signed observed Template length If all segments are mapped to the same reference the unsigned observed template length equals the number of bases from the leftmost mapped base to the rightmost mapped base The leftmost segment has a plus sign and the rightmost has a minus sign The sign of segments in the middle is undefined It is set as 0 for single segment template or when the information is unavailable SEQ Segment sequence This field can be a when the sequence is not stored If not a the length of the sequence must equal the sum of lengths of M I S X operations in CIGAR An denotes the base is identical to the reference base No assumptions can be made on the letter cases QUAL ASCII of base quality plus 33 same as the quality string in the Sanger FASTQ format A base quality is the phred scaled base error probability which equals 1010g 0 Pr This field can be a when quality is not stored If not a SEQ must not be a and the length of the quality string ought to equal the length of SEQ HD VN 1 5 SO coordinate SQ SN ref LN 45 r001 163 r002 0 r003 0 r004 0 r003 2064 r001 83 ref 7 30 8M214M1D3M TTAGATAAAGGAT ref 9 3S6M1P114M AAAAGATAAGGAT ref 9 5S6 GCCTAAGCTAA SA Z ref 29 6H5M 17 0 ref 16 6M14N5M ATAGCTTCAGC ref 29 6H5 TAGGC SA Z ref 9 5S6M 30 1 ref 37 9M CAGCGGCAT NM i 1 Figure 2 3 SAM file example with tabs replaced with spaces for readability The SAM BA
19. splice junction acceptor site is used changing the 5 boundary of the downstream exon Figure 4 3 Alternative 3 Figure 4 3 Alternative 3 alternative splicing Alternative 5 donor site An alternative 5 splice junction donor site is used changing the 3 boundary of the upstream exon Figure 4 4 36 Alternative Splicing Analysis Process Alternative 5 Figure 4 4 Alternative 5 alternative splicing Intron retention An intron can remain in the mature mRNA molecule or be spliced out This differs from exon skipping because the retained sequence is not flanked by introns If the retained intron is in the coding region the intron must encode amino acids in frame with the neighbouring exons or a stop codon or a shift in the reading frame will cause the protein to become non functional Figure 4 5 Intron retention Figure 4 5 Intron retention alternative splicing There are two more modes of alternative splicing although they are less frequent alternative first exon and alternative last exon One useful measure to extract is the count of each type of junction Many programs can achieve this They usually require an annotation file GTF and a file with mapped reads SAM BAM As explained in Chapter 2 if a genome has an error and misses it s ending frame it will end when it finds the next ending frame and this will generate a potentially harmful protein As such we can use open readi
20. the following frameworks installed on the operating system e NET framework 4 0 or newer e Mono 3 2 8 or newer The solution has two main projects Their prerequisites are Web service e MongoDB The web service requires a valid connection to a MongoDB database The MongoDB database doesn t need to be on the same computer Worker bot 6 Microsoft NET framework http www microsoft com net 6 Mono http mono project com Main_Page 6 MongoDB http www mongodb org 45 Web Service User Manual e A command line shell The worker bot uses a shell to run jobs locally On Windows cmd exe is used and is installed with the system On Linux Bourne shell sh is used and either that or a compatible shell must be installed e A list of packages The worker bot uses command line programs to run the jobs and these programs must be installed and be available to be run from the default shell Currently these are the required programs o Wget Wget is a program to download content from HTTP and FTP servers It is usually installed on Linux and can be installed on Windows using Gow o EMBOSS EMBOSS is a collection of command line programs for bioinformatics Some of those programs are used by some jobs o SAMtools SAMtools is used to work with SAM BAM files o InterProScan InterProScan in a web service that uses EMBOSS tools in the background to search for proteins in their databas
21. 3 chrY chr2 random or scaffold e g scaffold10671 CHROMSTART The starting position of the feature in the chromosome or scaffold The first base in a chromosome is numbered 0 2 BED format http www genome ucsc edu FAQ FAQformat html format 10 Basic Concepts and Survey on Technology 3 CHROMEND The ending position of the feature in the chromosome or scaffold The CHROMEND base is not included in the display of the feature For example the first 100 bases of a chromosome are defined as CHROMSTART 0 CHROMEND 100 and span the bases numbered 0 99 The 9 additional optional BED fields are 4 NAME Defines the name of the BED line This label is displayed to the left of the BED line in the Genome Browser window when the track is open to full display mode or directly to the left of the item in pack mode 5 SCORE A score between 0 and 1000 If the track line useScore attribute is set to 1 for this annotation data set the score value will determine the level of grey in which this feature is displayed higher numbers darker grey Figure 2 6 shows the Genome Browser s translation of BED score values into shades of grey lt 166 167 277 278 388 389 499 500 611 612 722 723 833 834 944 2945 Figure 2 6 Range of score values and it corresponding grey colour STRAND Defines the strand either or THICKSTART The starting position at which the feature is drawn thickly for example the start codon in gen
22. 9261174 doi 10 1186 gb 2009 10 3 r25 Quevillon E V Silventoinen S Pillai N Harte N Mulder R Apweiler and R Lopez 2005 InterProScan protein domains identifier Nucleic Acids Research no 33 suppl 2 W116 W120 http nar oxfordjournals org content 33 suppl_2 W116 abstract doi 10 1093 nar gki442 Roberts A H Pimentel C Trapnell and L Pachter 2011 Identification of novel transcripts in annotated genomes using RNA Seq Bioinformatics no 27 17 2325 2329 lt Go to ISI gt WOS 000294067300001 doi DOI 10 1093 bioinformatics btr355 Roberts A C Trapnell J Donaghey J L Rinn and L Pachter 2011 Improving RNA Seq expression estimates by correcting for fragment bias Genome Biol no 12 3 R22 http www ncbi nlm nih gov pubmed 21410973 doi 10 1186 gb 2011 12 3 122 Sette Claudio Michael Ladomery and Claudia Ghigna 2013 Alternative Splicing Role in Cancer Development and Progression International Journal of Cell Biology no 2013 2 http dx doi org 10 1155 2013 421606 doi 10 1155 2013 421606 Trapnell C D G Hendrickson M Sauvageau L Goff J L Rinn and L Pachter 2013 Differential analysis of gene regulation at transcript resolution with RNA seq Nature Biotechnology no 31 1 46 lt Go to ISI gt WOS 0003 13563600021 http www ncbi nlm nih gov pmc articles PMC3869392 pdf nihms439296 pdf doi Doi 10 1038 Nbt 2450 Trapnell C L Pachter and S L Salzberg 2009
23. A 3 6 1 Bam2Sam Converts a BAM file to the SAM format Download Downloads a file to a folder from an HTTP or FTP server InterProScan Uses the InterProScan sequence search web service to scan FASTA files for known proteins Orf Find Runs an ORF finder job using a FASTA file as input and outputs an ORF file Sam2Bam Converts a SAM file to the BAM format Sam2Fasta Converts a SAM file to the FASTA format Add Job You can add a job from the menu You need to enter the following fields to create a job Figure A 15 Oo Name Job name Status Job status from the following o Pending awaiting execution Use this to create a job but not have it executed right away o Ready ready for execution by a worker bot Use this to signal the job is to be executed when a worker bot is available Project Optional field associating a job to one of the user s projects This will run the job in that project s work area and create new files there instead of in the root folder You also have access to input files located inside that project s work area List of parameters Each job takes a list of parameters that vary from job type Usually it is an input file or more and a name of an output file 84 InterProScan web service https www ebi ac uk Tools webservices services pfa iprscan5 soap 59 Web Service User Manual Jobs Add Convert Bam to Sam Job Add Cancel Figure A 15 Add job examp
24. FACULDADE DE ENGENHARIA DA UNIVERSIDADE DO PORTO A Computational Platform for Assessing the Impact of Alternative Splicing in Cancer Vitor Amalio Maia Martins Moreira PORTO FEU FACULDADE DE ENGENHARIA UNIVERSIDADE DO PORTO Mestrado Integrado em Engenharia Informatica e Computa o Supervisor Rui Camacho Departamento de Engenharia Informatica da FEUP Second Supervisor Pedro Ferreira Swiss Institute of Bioinformatics June 23 2014 Vitor Am lio Maia Martins Moreira 2014 A Computational Platform for Assessing the Impact of Alternative Splicing in Cancer Vitor Amalio Maia Martins Moreira Mestrado Integrado em Engenharia Informatica e Computa o Aprovado em provas publicas pelo Juri Presidente Ana Paiva PhD Vogal Externo S rgio Matos PhD Orientador Rui Camacho PhD July 11 2014 Abstract Cancer is a disease that affects millions of individuals around the world every year Although it is known by this single name it may have different causes One such cause is based on an abnormal working of the genetic mechanisms The emergence of high throughput sequencing has brought substantial advances in cancer genomics research The current pace of advance of this technology made it possible to assay with high depth of coverage the DNA and the transcriptomes RNA Seq of tumour samples and cancer cell lines Aberrant RNA processing of genes plays an important role in cancer The goal of this thes
25. M format as seen in Figure 2 3 is a bit complex and is not fully explained in this document For a more detailed description see the format s specification GTF GTF Gene Transfer Format is a file format used to hold information about a gene and is the output format of differential expression software such as Cufflinks It is an extension of the GFF General Feature Format which has 9 tab separated required fields but with the last field different as seen in Figure 2 5 1 2 3 SEQNAME The name of the sequence Must be a chromosome or scaffold SOURCE The program that generated this feature FEATURE The name of this type of feature Some examples of standard feature mm types are CDS start codon stop codon and exon SAM format http samtools sourceforge net SAMv1 pdf 10 GTF format http www genome ucsc edu FAQ F AQ format html format4 GFF format http www genome ucsc edu FAQ F AQ format html format3 9 lt 166 AB000381 AB000381 AB000381 AB000381 AB000381 BED Basic Concepts and Survey on Technology START The starting position of the feature in the sequence The first base is numbered 1 END The ending position of the feature inclusive SCORE A score between O and 1000 If the track line USESCORE attribute is set to 1 for this annotation data set the score value will determine the level of grey in which this feature is displayed higher numbers
26. R operations are given in the following table set if not available Op BAM Description 0 Alignment match can be a sequence match or mismatch Insertion to the reference Deletion from the reference Skipped region from the reference Soft clipping clipped sequences present in SEQ Hard clipping clipped sequences NOT present in SEQ Padding silent deletion from padded reference vurvzo0o 2 X 7 RNEXT Reference sequence name of the primary alignment of the next read in the template For the last read the next read is the first read in the template If SQ header lines are present RNEXT if not or must be present in one of the SQ Sequence match Sequence mismatch CON DU A uN HR SN tag This field is set as when the information is unavailable and set as if RNEXT is identical RNAME If not and the next read in the template has one primary mapping see also bit 0x100 in FLAG this field is identical to RNAME at the primary line of the next read If RNEXT is no assumptions can be made on PNEXT and bit 0x20 8 PNEXT Position of the primary alignment of the NEXT read in the template Set as 0 when the information is unavailable This field equals POS at the primary line of the next read If PNEXT is 0 no assumptions can be made on RNEXT and bit 0x20 Phred quality score http en wikipedia org wiki Phred_quality score 8 10 11 Basic Concepts and Survey on Technology TLEN
27. Service contains the web service web logic including views and JavaScript Uses the logic layer to access the database WebService Common Contains common code between all projects such as settings code WebService Common Logic contains the code to interface with the database WebService Common WS contains the code to interface with the web service via the REST API WebService Self program that creates a self hosting web service instance WebService Aspnet program that creates a web service instance hosted under Microsoft s Internet Information Services IIS WebService WorkerBot program that uses the WS layer to access the web service WebService WorkerBot _ webservice Self Webservice Aspnet WebService Common Ws C vV Y V WebService WebService Common Logic WebService Common Figure 3 3 Web service project layers 4 Internet Information Services 27 A Website for Alternative Splicing Analysis There is a common set of functionalities that are used by both web service and worker bot projects These include utility functions for common operations such as handling settings files The logic layer interfaces with the database and provides an abstraction to manage the business logic The logic layer is used only by the web service layer which implements the user interface and a REST API to access it remotely The web service layer is used by both the Self a
28. Stdout Stderr transfer e File events where a job is started when a file is created e File transfer to transfer files from to the server e Authentication and role based security e User interfaces 38 EMBOSS http emboss sourceforge net 39 EMBOSS interfaces http emboss sourceforge net interfaces 4 EMBOSS workflows http emboss sourceforge net interfaces workflows Job scheduling solutions http en wikipedia org wiki List_of job scheduler software 22 Basic Concepts and Survey on Technology 2 7 Chapter Conclusions In this chapter we outlined some of the tools and algorithms we will use to perform our work Some of the algorithms are best suited for large amounts of data while others are not The biological analysis tools are not thoroughly mentioned since they will not all be used and the ones mentioned are most likely to be chosen as they are usually paired together Because the focus will be on the computational platform the tools are not that important One advantage of using the tools mentioned is that they were made to work well with each other and were partially developed by the same people Regarding the algorithms tools they themselves pose challenges related to data integration The algorithms themselves can have different behaviour depending on the size of the training sets and the amount of data to analyse 23 Chapter 3 A Website for Alternative Splicing Analysis Most of the a
29. a in the BAM file to the BED format It takes an infile bam as input and outputs an outfile bed After this step a command is ran to cross reference this file with a GTF file and count the gene status column This will give us a count of the known and unknown genes KNOWN 548118 NOVEL 1626827 PUTATIVE 1532709 Another very important step in the analysis process is the identification of open reading frames ORFs are the regions of the nucleotide sequence from the start codon to the stop codon Gene finding is usually started by searching for open reading frames An ORF is a sequence of DNA that starts with start codon ATG not always and ends with any of the three termination codons TAA TAG TGA There is a task in the platform to execute this as well 6 GENCODE GTF format http www gencodegenes org gencodeformat html 38 Alternative Splicing Analysis Process 4 2 Search for Known Proteins Domains A useful analysis that can be performed is searching for known proteins domains The interest of the last step relies on the possibility of this information to be quite relevant to explain diseases When there is an aberrant splicing the translation of DNA into RNA might be stopped too early which may have the consequence that the coded protein will be too short and some very important domains of such protein will be missing This is done using InterProScan Zdobnov and Apweiler 2001 Quevillon et al 2005 a web serv
30. aa dans pad 2 1 4 Structure of the Thesis 3 2 3 Basic Concepts and Survey on Technology cccscsscsscssscecsccsesccesesscrescccesecees 4 2 1 Basic Biological Concepts ooooocococonocononanonnnonnnonnnnnnonan nono nono ncon ccoo noo 4 2 2 Software for the Biological Analysis ooooonnccnnncniocononcnonoconoconoconocnns 6 2 2 1 Standard File Formats for Bioinformatics Data oooonccnicnninncinn o 7 2 2 2 Genomic Assemblage Software c cccsccescesecesecseeeeeeeeeseeseeesaes 13 2 2 3 Databases llas ltda niet en a 14 2 3 Data Mant ai tn mr ete RUE Es 15 2 3 1 Data Analysis Algorithms ooooooocnnonnconoocnoonncnnncnnncon cnn nono noconocnnnnos 15 2 3 2 Data Analysis Tools ooonoconnoninonoconoconoocnnonncnn con noon conocio a 16 2 4 Algorithm Evaluation Methods and Measures 18 2 5 WEDISELVICES td iaa lili 19 2 5 1 Representational State Transfer 20 2 5 2 Web Frameworks annur arane e a N 21 2 6 JOD Scheduling matias iio 21 2 7 Chapter Conclusions sias st a A EA EA 23 3 A Website for Alternative Splicing Analysis ss seseeseeese 24 3 1 Web SERVICO ci eset Me aie nn tete Minnie 24 3 1 1 Main Use Cas ati 24 3 2 Architects das 26 32 1 Database Architecture es 28 3 2 2 Web Service Architecture ccccccssccssecsseceneceseceeeeseeeeeeeseeeesseennes 30 3 3 Chapter Conclusions 4 34 4 Alternative Splicing Analysis Process s
31. arning models with associated learning algorithms that analyse data and recognize patterns used for classification and regression analysis Given a set of training examples each marked as belonging to one of two categories in the case of classification an SVM training algorithm builds a model that assigns new examples into one category or the other making it a non probabilistic binary linear classifier An SVM model is a representation of the examples as points in space mapped so that the examples of the separate categories are divided by a clear gap that is as wide as possible New examples are then mapped into that same space and predicted to belong to a category based on which side of the gap they fall on In addition to performing linear classification SVMs can efficiently perform a non linear classification using what is called the kernel trick implicitly mapping their inputs into high dimensional feature spaces 21 Wikipedia is a free encyclopaedia http en wikipedia org 15 Basic Concepts and Survey on Technology Ensembles In statistics and machine learning ensemble methods use multiple models to obtain better predictive performance than could be obtained from any of the constituent models Unlike a statistical ensemble in statistical mechanics which is usually infinite a machine learning ensemble refers only to a concrete finite set of alternative models but typically allows for much more flexible structure to exist
32. at pieces of RNA are collected RNA sequencing out of small sequences These short RNA sequences are commonly referred to as reads Genome sequencing is the process of mapping out the order of the genome s bases that make up an organism s DNA By knowing which genes do what we can map an organism s characteristics to their equivalent genes and thus know which are responsible for what parts of that organism When applied to diseases this allows us to know which genes are responsible for their development The human genome is made up of over 3 billion of these base molecules and we are not one of the most complex species on earth as other species possess far more molecules In order to obtain the whole human genome one first splits it into pieces which are then sequenced and reassembled in the original order to get to the original genome As you can imagine this process is error prone because of incorrect reads from the machinery used and the repetition and mapping of the reads in their right order The computer algorithms used to perform this have varying results depending on the species as some species have smaller strains with many repetitions where others have wider more unique strains Many techniques are used to perform genome sequencing Some are faster than others at the cost of reliability It is like solving a jigsaw puzzle where many of the puzzle pieces are very similar however this process isn t necessary to understand for the success
33. b based interface that makes it easy for the user to automate processes requiring the execution of several tasks e A software tool to be used in a collaborative process among researchers sharing any information they think useful to others e A ubiquitous computational tool accessible from any place with access to the web Introduction Contributions specific for the study of aberrant alternative splicing include e A platform running standard tools used in aberrant alternative splicing studies e The possibility of chaining tools to build a pipeline that accepts the aligned reads as input and produces the aberrant alternative splicing results e An automatic process using the API from the adequate web resource to fetch the final information of the analysis concerning the domains of missing parts of the proteins encoded by the gene under analysis 1 4 Structure of the Thesis The rest of the thesis is structured as follows Chapter 2 describes the biological processes related to the domain of genetic based cancer and molecular biology basic concepts We also survey the main technologies that we have used in the development of the proposed computational platform Chapter 3 details the computational platform developed and how it works both internally and from an end user s perspective Chapter 4 describes the biological processes important for alternative splicing analysis and how they can be executed using our proposed computational pla
34. be able to add a description to a file describing it in more detail e The solution must enforce edit restrictions A user should not be able to alter or delete another user s artefacts unless he is an administrator e The solution must use open source technologies The solution should not use technologies with restrictive licenses that prevent usage in commercial applications or require the purchase of a license to use 2 Mouse hover http en wikipedia org wiki Mouseover 25 A Website for Alternative Splicing Analysis Web Service C Create User Account O Manage Files ser Manage Jobs extends Manage Projects Guest extends Manage Clusters O Validate New User Accounts extengs Administrator Manage Users Figure 3 1 Web service use cases 3 2 Architecture The web service s high level architecture consists of a server that manages the logic and a number of worker bots Clusters that are idle and ping the web service for pending jobs as seen in Figure 3 2 Worker bots download all necessary files from the server before running a job and then upload back the results This allows the service to have many worker bots executing tasks asynchronously 26 A Website for Alternative Splicing Analysis Web service Cluster 2 User 1 User 2 Figure 3 2 Web service architecture Internally the web service is split into the following projects Figure 3 3 Web
35. between those alternatives Random Forrest Random forests are an ensemble learning method for classification and regression that operate by constructing a multitude of decision trees at training time and outputting the class that is the mode of the classes output by individual trees K NN K Nearest neighbours algorithm or k NN for short is a non parametric method used for classification and regression In both cases the input consists of the k closest training examples in the feature space The output depends on whether k NN is used for classification or regression Inductive Logic Programming Inductive logic programming ILP is a sub field of machine learning which uses logic programming as a uniform representation for examples background knowledge and hypotheses Given an encoding of the known background knowledge and a set of examples represented as a logical database of facts an ILP system will derive a hypothesised logic program which entails all the positive and none of the negative examples This is a good candidate to start with as it has been used in these types of problems mainly because of its ease of handling multi relational data ILP can build complex but comprehensible models that make it easy to explain the phenomena that produced them 2 3 2 Data Analysis Tools Here we will describe packages used for the data analysis part Some of these tools are used by other tools CummeRbund 16 Basic Concepts an
36. cell lines K562 and GM12878 The very first step in the whole process is to download the data When using the platform to downloading the K562 cell line CSHL1 took 1 hour 50 minutes and 13 GB were downloaded For GM12878 cell line CSHL2 1 hour and 54 minutes was required to download 15 GB of data the BAM file with the aligned reads 4 1 Junctions Identification and Open Reading Frame Detection Alternative splicing is a process where during gene expression a particular exon may be included or excluded from the final processed messenger RNA mRNA from that gene The proteins translated from the alternative spliced mRNA will have differences in their structure as explained in Chapter 2 The platform was used to identify the following five basic modes of alternative splicing The ENCODE project http www genome gov 10005107 35 Alternative Splicing Analysis Process Exon skipping also known as cassette exon where the exon can be spliced out of the transcript or retained with its flanking introns This is the most common mode of alternative splicing in mammals Figure 4 1 Exon skipping Figure 4 1 Exon skipping alternative splicing Mutually exclusive exons One of two exons is retained in mRNAs after splicing but not both in a mutually exclusive manner Figure 4 2 Mutually exclusive exons Figure 4 2 Mutually exclusive exons alternative splicing Alternative 3 acceptor site An alternative 3
37. d Survey on Technology CummeRbund is an R package that is designed to aid and simplify the task of analysing Cufflinks RNA Seq output R is program and programming language used in statistical analysis RapidMiner RapidMiner is a software platform developed by the company of the same name that provides an integrated environment for machine learning data mining text mining predictive analytics and business analytics It is used for business and industrial applications as well as for research education training rapid prototyping and application development and supports all steps of the data mining process including results visualization validation and optimization RapidMiner is developed on a business source model which means the core and earlier versions of the software are available under an OSl certified open source license A Starter Edition is available for free download Personal Professional and Enterprise Editions are available KMINE KNIME the Konstanz Information Miner is an open source data analytics reporting and integration platform KNIME integrates various components for machine learning and data mining through its modular data pipelining concept A graphical user interface allows assembly of nodes for data pre processing ETL Extraction Transformation Loading for modelling and data analysis and visualization Since 2006 KNIME has been used in pharmaceutical research but is also used in other areas li
38. darker grey If there is no score value enter Figure 2 4 shows the Genome Browser s translation of BED score values into 167 277 278 388 389 499 500 611 612 722 723 833 834 944 2945 shades of grey Figure 2 4 Range of score values and it corresponding grey colour STRAND Valid entries include or for don t know don t care FRAME If the feature is a coding exon frame should be a number between 0 2 that represents the reading frame of the first base If the feature is not a coding exon the value should be ATTRIBUTE A list of tag value pairs providing additional information about each feature Attributes must end in a semi colon and be separated from any following attribute by exactly one space Twinscan CDS 380 ar E 001 transcript id Twinscan CDS 501 a t 001 transcript id winscan CDS 700 o 001 transcript id winscan start codon 380 E 001 transcript_id winscan stop codon 708 se SF 001 transcript id Figure 2 5 GTF file example with tabs replaced with spaces for readability BED is a file format used to hold sequence annotations It has 3 mandatory fields and 9 optional ones as seen in Figure 2 7 The order of the optional fields must be preserved where previous fields are not left empty This format is sometimes called BED6 or BED12 depending on the number of fields it has The first three required BED fields are CHROM The name of the chromosome e g chr
39. donor with northern and western European ancestry by EBV transformation The mapped data can be obtained from here http genome crg es encode RNA dashboard hg19 18 SEQAnswers wiki http seqanswers com wiki 12 The ENCODE project http www genome gov 10005107 2 The ENCODE project common cell types http www genome gov 26524238 14 Basic Concepts and Survey on Technology 2 3 Data Mining Here we present the algorithms that are being considered for the data analysis phase and software packages that can be used to apply them Descriptions are taken from Wikipedia and or their web pages for the tools 2 3 1 Data Analysis Algorithms Decision Trees A decision tree is a flowchart like structure in which internal node represents test on an attribute each branch represents outcome of test and each leaf node represents class label decision taken after computing all attributes A path from root to leaf represents classification tules Rule Induction Algorithms Rule induction algorithms are an area of machine learning in which formal rules are extracted from a set of observations The rules extracted may represent a full scientific model of the data or merely represent local patterns in the data Many branches of machine learning apply this technique namely inductive logic programming Support Vector Machines Support vector machines SVMs also known as support vector networks are supervised le
40. e 3 7 Figure 3 8 Figure 4 1 Figure 4 2 Figure 4 3 Figure 4 4 Figure 4 5 Figure A 1 Figure A 2 Figure A 3 Figure A 4 Figure A 5 Figure A 6 Figure A 7 Principles of alternative splicing SAM file example with tabs replaced with spaces for readability Range of score values and it corresponding grey colour GTF file example with tabs replaced with spaces for readability Range of score values and it corresponding grey colour BED file example with tabs replaced with spaces for readability FASTA file example Web service use cases Web service architecture Web service project layers Database schema of the web service Web service initial menu BAM to SAM conversion job example Markdown example File browser Exon skipping alternative splicing Mutually exclusive exons alternative splicing Alternative 3 alternative splicing Alternative 5 alternative splicing Intron retention alternative splicing Web service example configuration file Worker bot example configuration file Sign in form User account menu Login menu Web service menu Cluster accounts listing 1x 10 10 11 12 13 26 27 27 29 30 31 32 33 36 36 36 37 37 49 51 52 53 53 54 55 Figure A 8 Add cluster account Figure A 9 File browser Figure A 10 Figure A 11 Figure A 12 Figure A 13 Figure A 14 Figure A 15 Figure A 16 Figure A 17 Figure A 18 Figure A 19 Figure A 20 Edit file
41. e displays 8 THICKEND The ending position at which the feature is drawn thickly for example the stop codon in gene displays 9 ITEMRGB An RGB value of the form RGB e g 255 0 0 If the track line ITEMRGB attribute is set to On this RBG value will determine the display colour of the data contained in this BED line NOTE It is recommended that a simple colour scheme eight colours or less be used with this attribute to avoid overwhelming the colour resources of the Genome Browser and your Internet browser 10 BLOCKCOUNT The number of blocks exons in the BED line 11 BLOCKSIZES A comma separated list of the block sizes The number of items in this list should correspond to BLOCKCOUNT 12 BLOCKSTARTS A comma separated list of block starts All of the BLOCKSTART positions should be calculated relative to CHROMSTART The number of items in this list should correspond to BLOCKCOUNT Track definition lines can be used to configure the display further e g by grouping features into separate tracks Track lines should be placed at the beginning of the list of features they are to affect The track line consists of the word track followed by space separated key value pairs see the example below Valid parameters used by Ensembl are 11 Basic Concepts and Survey on Technology NAME unique name to identify this track when parsing the file DESCRIPTION Label to be displayed under the track in Region in Deta
42. e from FASTA sequences The web service uses the C client IPRScan5CliClient which can be found on the web service s web page On Linux based systems most of these programs can be installed easily using the default package manager On Debian based operating systems you can use the following commands to install all the necessary dependencies requires root access apt get install mongodb apt get install mono runtime apt get install emboss apt get install samtools apt get install wget To install the InterProScan client you need to download the C client IPRScan5CliClient exe to a folder that will be available from any command line prompt On a Debian based system this folder can be usr bin After that make sure the program can be run without the exe extension Simply rename the file to remove the extension Mono must be installed to run the program On Windows you do not need to rename the program but simply make it available from the command line prompt A 2 2 Compiling The root of the solution should look like this 6 Bourne shell http en wikipedia org wiki Bourne shell 6 Wget http www gnu org software weget Gnu on Windows http github com bmatzelle gow wiki 7 EMBOSS http emboss sourceforge net 71 SAMtools http samtools sourceforge net 72 InterProScan web service http www ebi ac uk Tools webservices services pfa iprscan5_soap 73 Debian operating system http www debian org
43. een 35 4 1 Junctions Identification and Open Reading Frame Detection 35 vil 4 1 1 Obtaining the Junctions Type Count 4 2 Search for Known Proteins Domains oooccoconinnnnocnnonnnonnconncon conoces 4 3 Other iss re mnt en A ie COTCIUSIONS si csiccecssscsesscssvsdscssaccaunctscbnsesdesssenessupiseedosssenassnbasessbsssepesccedsestoosbepasscunas 5 1 Achievement of Objectives ooooooocccoconococoncconaconononnnnonanannonn cono ccnn noo 5 2 Future Work E Web Service User Manual ssesssesssessoessocsssesssesssesssesssesssesssessseossoessosssoossoossoosso Al TEMO UCUON PARADO RE DORIAN SS eh read cb EA DE A 2 Installation dia adie A 2 1 PHETEQUISITES tias be pasto AO A 2 2 COMPUTO Fritos Ai tit rs hea vs far rte A 2 3 A REN A 2 3 1 Web Service A 2 3 2 Worker Bot A 3 Us AGUS Re da anda Ne Are A 3 1 SIA A Sikes nt Me A 3 2 RA OS A 3 3 Web Service Me iii ita tail A34 Cluster a Pre Pr a en eee anda A 3 4 1 Create Cluster A 3 5 Pl A ia da A 3 6 JOOS ds bed A 3 6 1 Add Job A 3 6 2 Job Details A 3 7 POS Aaa A 3 7 1 Create Project A 3 7 2 Project Details A 3 8 A tate cee Rt Nm ee AR EE ir ent A 3 8 1 Pending User viii List of Figures Figure 2 1 Diagram showing the translation of mRNA and the synthesis of proteins by a ribosome Figure 2 2 Figure 2 3 Figure 2 4 Figure 2 5 Figure 2 6 Figure 2 7 Figure 2 8 Figure 3 1 Figure 3 2 Figure 3 3 Figure 3 4 Figure 3 5 Figure 3 6 Figur
44. efault theme is a custom theme based on the ones found in bootswatch com by Thomas Park You can alter the theme by replacing the files bootstrap css and bootstrap min css located in the Content folder with new ones Note however that the fonts file path should start with fonts and not fonts Most themes haven t got this path correctly and must be manually changed 81 SHA256 hashing algorithm http en wikipedia org wiki SHA 2 82 SHAI hashing algorithm http en wikipedia org wiki SHA 1 Twitter Bootstrap http getbootstrap com 50 Web Service User Manual A 2 3 2 Worker Bot WebService WorkerBot ex WebService WorkerBot exe config The worker bot supports the h or help parameters that print the application s supported configuration parameters and their default values URL URL pointing to the web service The HTTP protocol string must be present otherwise an exception will be thrown Default is http localhost 8080 Login Name of the cluster account to use to authenticate This account must exist on the web service this worker bot is connecting to PassHash Hash of the password used to authenticate using the SHAI hashing algorithm LocalBot Boolean value indicating if bot is sharing the data and the stdio folders with the web service in the same computer This prevents the worker bot from trying to download remote files and upload the results back Default is false DataFolder
45. ent it wants in an easy way Whether it is an HTML page or a JSON XML response Many web frameworks provide mechanisms to deal with these issues and other common ones For the solution we are developing Routing model binding content negotiation and view engine support are the most important ones 2 6 Job Scheduling Since most bioinformatics applications are command line based and abide by the Unix philosophy of having programs chained together to produce a final result from a set of inputs the web service that will manage the data analysis will be a job scheduling application where jobs are created by the user to perform a specific task and ran in a worker bot that will be idle until a task is given to it Because job scheduling is a problem that has been addressed many times before there are already many solutions available There are also some solutions made specifically for bioinformatics Sequence Manipulation Suite 36 List of web application frameworks http en wikipedia org wiki Comparison_of web application frameworks 37 Job Scheduling http en wikipedia org wiki Job scheduler 21 Basic Concepts and Survey on Technology Sequence Manipulation Suite is a collection of JavaScript programs for generating formatting and analysing short DNA and protein sequences It is commonly used by molecular biologists for teaching and for program and algorithm testing This suite does not offer user management or a w
46. eters a worker bot can execute the job using whatever tools there are on the platform and correct and improve the command pipeline to execute for each action without needing to alter the job For instance to execute a conversion job we can use one program and later on change that program to a different version as long as the job s parameters are used in the same way This is useful when upgrading the pipeline with a different program to test performance or accuracy as a worker bot can use a newer version than another bot Once a job is completed the program s stdout and stderr output are saved and synchronized with the service so the user can see their output if any The worker bots poll the web service every 10 seconds for a pending job We can monitor the status of the worker bots by viewing the clusters page and see the last polling time Projects exist to allow the user to describe the project and the files associated with it Projects have a description which uses markdown Figure 3 7 as a text description to allow for richer descriptions Project Title Project Title Paragraphs are separated by a blank line Text attributes italic bold monospace A link http example com Testing files genome reference gtf testing set bam 1 1 Introduction Paragraphs are separated by a blank line Text attributes italic bold monospace A link Testing files e genome reference gtf e testing set bam
47. f data can have One of these metrics is named F 27 R http www r project org 8 Overfitting http en wikipedia org wiki Overfitting 2 Cross Validation http en wikipedia org wiki Cross validation statistics 39 Resampling http en wikipedia org wiki Resampling statistics 18 Basic Concepts and Survey on Technology measure which takes into account both the precision and the recall of the test to compute the final score 2 5 Web Services Web services are a technology that works over computer networks by providing a means of communication between machines using a communication interface that is understood by all machines This is usually implemented on top of HTTP Hypertext transfer protocol protocol used to access the internet from a web browser in conjunction with other web standards and formats The usefulness of web services is that they sit on top of well established and standard protocols that are widely available on nearly all computer systems and thus can have a broader reach The HTTP protocol works by requesting an operation pointed at a hyperlink that will in turn reply with an appropriate response For example when we navigate on a web browser to Google s search engine hyperlink http www google com we are performing the GET request method on that hyperlink and will get Google s initial web page as a response There are other request methods we can specify that allow us to pe
48. he execution of the bioinformatics jobs was completed It is multi platform having been tested in both Windows and Linux The web framework is divided into 2 main parts the web service itself and the worker bot that uses the web service to run tasks from another computer The web service meets all the requirements set in terms of functionality namely being multi platform having a user interface that works well with small screens and touch screens user accounts and authentication management of files management of projects describing experiences management of jobs to be run on worker bots being able to run jobs from another computer taking care of file transfers between both computers automatically Jobs in the web service are generic enough to allow the creation of new types of jobs without having to rewrite the underlying logic by abstracting away command line programs and storing only the name of the command or action to perform and a list or arguments to pass to the command This allows us to maintain backwards compatibility by ignoring newer arguments and change the programs used in a job without having to alter existing jobs as long as the new command interprets the arguments in the same order as before In summary the thesis work provides a general purpose architecture to run jobs in several machines running different operating systems makes available to the community a web based interface that makes it easy for any user to automate proce
49. ications on all platforms It is not recommended to mix and match binaries compiled with both versions Each project is compiled to each of the project s bin folder bin Debug or bin Release After building the project all the necessary dependencies for the web service project should be in the WebService Self bin folder e Content e Views 74 Visual Studio express editions http Avww microsoft com express download 7 MonoDevelop http monodevelop com 76 NuGet package manager http www nuget org 77 MSBuild http msdn microsoft com en us library dd393574 aspx 78 XBuild http www mono project com Microsoft Build 47 Web Service User Manual MongoDB Bson dll MongoDB Driver dll Nancy Authentication Forms dll Nancy dll Nancy Hosting Self dll Nancy ViewEngines Razor dll System Web Razor Unofficial dll WebService Common dll WebService Common Logic dll WebService dll WebService Self ex WebService Self exe config The Content folder contains static web content files namely CSS JavaScript and static HTML pages The Views folder contains the Razor view engine s views files cshtml that render the HTML views at runtime The other project in the solution is the worker bot All the necessary dependencies should be in the WebService WorkerBot bin folder Newtonsoft Json dll RestSharp dll WebService Common dll WebService Common WS dll WebService WorkerBot ex
50. ice for automated sequence analysis of proteins that can identify regions of interest based on the InterPro consortium member databases This allows us to quickly detect a novel sequence with considerable confidence There are two versions available a standalone version and a web service We use the web service with the following command C version of the web service client IPRScan5CliClient mail email example com sequence infile fasta outfile outfilename The web service version requires a valid email address and a FASTA file as input with the nucleotide sequences to scan This tool can be used from a BAM file by converting it to a FASTA file using our web service 4 3 Other Other steps important for data analysis involve conversion between standard formats The following commands are used to convert between formats BAM to SAM using SAMtools samtools view h o outfile sam infile bam SAM to BAM using SAMtools samtools view bS infile sam gt outfile bam SAM to FASTA using EMBOSS segret infile sam outfile bam 62 InterProScan web service http www ebi ac uk Tools webservices services pfa iprscan5_soap 39 Chapter 5 Conclusions This chapter describes the solution developed regarding goals achieved and the contributions that were made Some suggestions for improvements are also made 5 1 Achievement of Objectives During this thesis the development of the web service to automatize t
51. iki JSON XML http en wikipedia org wiki XML 19 Basic Concepts and Survey on Technology 2 5 1 Representational State Transfer One of such web service or web API standards is representational state transfer REST It differs from other web service standards by being simpler and relying on a small subset of the HTTP protocol It is also not a protocol definition but an architectural style of achieving common operations This means we are not forced to implement the web service in a specified way but have the liberty to deviate from the reference style to accommodate our solution The REST protocol works based on these simple aspects e A base URL such as http www example com resource e Standard HTTP methods usually GET PUT POST and DELETE Here are two examples Tables 2 1 and 2 2 of how the management of users can look like using REST URL http www example com users GET Gets a list with all the users PUT Not usually used POST Adds a new user to the collection DELETE Deletes all users Table 2 1 REST operations for user collection URL http www example com user Item1 GET Gets the user that corresponds to Item1 PUT Updates the user s attributes POST Not usually used DELETE Deletes this user Table 2 2 REST operations for user item Because REST is an architectural style we are not forced to implement these specific operations using these HTTP methods One common modification is
52. il PRIORITY integer defining the order in which to display tracks if multiple tracks are defined 4 USESCORE a value from 1 to 4 which determines how scored data will be displayed Additional parameters may be needed as described below e Tiling array e Colour gradient defaults to Yellow Green Blue with 20 colour grades Optionally you can specify the colours for the gradient cgColourl cgColour2 cgColour3 as either RGB hex or X11 colour names and the number of colour grades cgGrades e Histogram e Wiggle plot 5 ITEMRGB if set to on case insensitive the individual RGB values defined in tracks will be used track name pairedReads description Clone Paired Reads useScore 1 chr22 1000 5000 cloneA 960 1000 5000 0 2 567 488 0 3512 chr22 2000 6000 cloneB 900 2000 6000 0 2 433 399 0 3601 Figure 2 7 BED file example with tabs replaced with spaces for readability FASTA FASTA is a text based file format used to hold nucleotide or amino acid sequences which are represented using single letter codes A sequence in FASTA format begins with a single line description followed by lines of sequence data as seen in Figure 2 8 The definition line defline is distinguished from the sequence data by a greater than gt symbol at the beginning The word following the gt symbol is the identifier of the sequence and the rest of the line is the description optional Normally identifiers are simply
53. is is twofold automate as much as possible the analysis process of genetic data for cancer studies and develop a computational platform that makes the powerful computational resources required for such analysis easily manageable and in a scalable fashion With these two goals in mind we have implemented part of a pipeline that performs the analysis from the RNA Seq data the set of reads to the determination of the possible protein domains that may be in the origin of the disease The platform also allows the enrichment of the analysis by searching for information in relevant databases available on the internet We have also developed it in a way that hides the computational resources from the user making it easy to use by the life sciences experts The platform is also very easy to manage enabling the use updating removing of resources on different operating systems and without impact to the analysis process It also enables the execution of the complete set or individual steps of the analysis pipeline Although the platform is general purpose the current version is tuned for to the application of methods that use high throughput information from the transcriptome of cancer samples to identify events of aberrant splicing and their impact in the transcript structure The developed platform was tested in case studies using different cell lines from databases available on the ENCODE project web site 11 Resumo O cancro uma doen a que afeta mi
54. izations There are some publically available genome browsers online that allow us to visualize and retrieve genomic information from a number of species Some of the most well known projects are e The Ensembl genome database project e The UCSC University of California Santa Cruz genome browser 2 2 1 Standard File Formats for Bioinformatics Data Many formats exist to represent the various stages of this process One of the most common formats is the SAM Sequence Alignment Map format and its binary equivalent BAM The following format descriptions are taken from either Wikipedia or a web page describing the format from the format s web page or a public format description from a renowned institution SAM BAM The SAM Sequence Alignment Map format and its binary equivalent BAM are the most common formats used for sequence data It is used by many bioinformatics tools Both format store the same information The SAM format is a TAB delimited text format consisting of a header section which is optional and an alignment section If present the header must be prior to the alignments Header lines start with while alignment lines do not Each alignment line has 11 mandatory fields for essential alignment information such as mapping position and a variable number of optional fields for flexible or aligner specific information Each alignment line has 11 mandatory fields List of genome browsers http e
55. ke customer data analysis business intelligence and financial data analysis WEKA Weka Waikato Environment for Knowledge Analysis is a popular suite of machine learning software written in Java developed at the University of Waikato New Zealand Weka is free software available under the GNU General Public License Orange 26 a ad Orange is a component based data mining and machine learning software suite featuring a visual programming front end for explorative data analysis and visualization and Python 22 CummeRbund http compbio mit edu cummeRbund 2 RapidMiner http rapidminer com 2 KNIME http www knime org 2 Weka http www cs waikato ac nz ml weka 26 Orange http orange biolab si 17 Basic Concepts and Survey on Technology bindings and libraries for scripting It includes a set of components for data preprocessing feature scoring and filtering modelling model evaluation and exploration techniques It is implemented in C and Python Its graphical user interface builds upon the cross platform Qt framework Orange is freely distributed under the GPL It is maintained and developed at the Bioinformatics Laboratory of the Faculty of Computer and Information Science University of Ljubljana Slovenia R R is a free software programming language and software environment for statistical computing and graphics The R language is widely used among statisticians and
56. ke the whole process complete the steps that compute the gene expression from raw reads need to be incorporated 6 Testing Nancy applications http github com NancyFx Nancy wiki Testing your application 41 Conclusions It is quite common in molecular biology to have tasks that are frequently done and are composed by a set of steps a pipeline A nice extension of the proposed web interface is an interface where a computer scientist could assemble easily such type of pipelines He she would upload or indicate the software to be used if already available scripts to convert between file formats if necessary and the pipeline would be stored as a single software tool Some steps in the pipeline could even get information from web data bases to be used in the pipeline analysis relieving the user from collecting some of the useful information from several web sites for example reference genomes As users biologists would only need to provide the input file s and get the results without the need to call several programs and with the need to know the inner working of the process This work is partially done however it can be improved further by making it easier to build job pipelines We would also be very interested in a near future to use Data Mining tools to help in the analysis process Tasks like the use of clustering as a tool for outliers detection could be useful for identifying and explaining junctions with an origin in aberrant
57. le 60 Web Service User Manual A 3 6 2 Job Details Once the job is executed it will change its state depending on whether it is executing or it has already finished with success or because of an error Figure A 16 e Executing Job is being executed by a worker bot e Completed Job execution has completed without errors e Error Job execution was halted because of an error This usually means the command line ran in the worker bot encountered a problem which can be due to an incorrect parameter or another unforeseen error If the job ran with errors but successfully the state should be completed Job Details Name Status Creation Date sam2bam completed 6 19 2014 1 07 58 PM Jobs Edit Delete Figure A 16 Job details The standard output and standard error streams are saved so you can check a job s success Once you stop needing these files you can delete the using the small delete button next to the output label You can also edit a job Usually you do this to change its state Manually editing job parameters is not recommended Deleting jobs can be done after the job ran and you don t need it anymore Stdio files related to that job are also deleted All files produced by that job are not so you can use them in other jobs 61 Web Service User Manual A 3 7 Projects Projects are descriptions of experiments a user wants to have for reference Each project has its own working folder so users can have
58. lh es de pessoas em todo o mundo todos os anos Apesar de ser conhecida por este nome pode ter diferentes causas Uma dessas causas baseada num mau funcionamento dos mecanismos gen ticos O surgimento do sequenciamento de alto rendimento trouxe avan os substanciais na rea de investiga o da gen mica do cancro Com o passo atual de desenvolvimento desta tecnol gica poss vel determinar com grande cobertura o DNA e os transcriptomas RNA Seq de exemplos c lulas cancer genas e de linhas celulares O processamento aberrante de genes tem um papel importante no cancro O objetivo desta tese duplo automatizar o quanto poss vel o processo de an lise de dados gen ticos para o estudo do cancro e desenvolver uma plataforma computacional que permita o uso de poderosos recursos computacionais para estudos que envolvam an lise de informa o de alto rendimento do transcriptoma de amostras cancer genas de forma f cil e escal vel Com estes dois objetivos em mente implement mos parte das sequ ncias de tarefas que executam a an lise a partir dos dados RNA Seq o conjunto de reads at determina o dos poss veis dom nios de prote nas que possam estar na origem da doen a A plataforma permite o enriquecimento da an lise ao procurar por informa o em bases de dados relevantes dispon veis na internet A plataforma foi desenvolvida de maneira a esconder os recursos computacionais do utilizador fazendo com que esta seja simples de
59. lternative Splicing Analysis Projects PK FK1 name passhash createdate isadmin iscluster pending status commandname args createdate ownerid projectid description createdate ownerid Figure 3 4 Database schema of the web service All tables have a generated id named _id which is a 24 char unique identifier composed of the current time machine id process id and random data Here is a description of each field Files Jobs Projects Name File s path that acts as a unique identifier for a file Description A string with the file s description Name Job s name Status Current status of the job New jobs can be pending or ready CommandName Name of the command to execute Args List of arguments to pass to the command CreateDate Timestamp with the creation date Ownerld Id of the user who created the job ProjectId Optional id of a project to associate the job with Name Project s name Description A string with the project s description CreateDate Timestamp with the creation date Ownerld Id of the user who created the project 6 MongoDB Objectld http docs mongodb org manual reference object id 29 A Website for Alternative Splicing Analysis Users e Login string used to uniquely identify a user and sign in into the web service e Name User s name e PassHash User s password in hashed form e C
60. n wikipedia org wiki Genome_browser Ensembl http Avww ensembl org UCSC Genome Browser http genome ucsc edu Wikipedia is a free encyclopaedia http en wikipedia org SAMtools http samtools sourceforge net 7 SAM format http samtools sourceforge net SAMv1 pdf 7 Basic Concepts and Survey on Technology 1 QNAME Query template name Reads segments having identical QNAME are regarded as coming from the same template A QNAME indicates the information is unavailable 2 FLAG Reference sequence NAME of the alignment See the full specification for a detailed description 3 RNAME Reference sequence NAME of the alignment If SQ header lines are present RNAME if not must be present in one of the SQ SN tag An unmapped segment without coordinate has a at this field However an unmapped segment may also have an ordinary coordinate such that it can be placed at a desired position after sorting If RNAME is no assumptions can be made about POS and CIGAR 4 POS 1 based leftmost mapping position of the first matching base The first base in a reference sequence has coordinate 1 POS is set as 0 for an unmapped read without coordinate If POS is 0 no assumptions can be made about RNAME and CIGAR 5 MAPQ Mapping quality It equals 10log Pr 8 rounded to the nearest integer A value 255 indicates that the mapping quality is not available 6 CIGAR CIGAR string The CIGA
61. nd Aspnet projects which are stub projects with minimum functionality that create an instance of the web service layer in different contexts a self hosting application or an IIS hosted one The common WS layer short for web service consumes the web service using the provided REST API Finally the worker bot uses the common WS layer to access the web service without having to know the underlying implementation The web service implements the following concepts e Clusters Accounts used by the worker bots to login into the service e Files Files stored in the web service s server e Jobs Job management for creating different types of jobs to be run by a worker bot e Projects Project describing an experiment Jobs can be associated to a project for easier management e Users User accounts for authentication into the system and authorization of operations 3 2 1 Database Architecture The database that supports the web service is a MongoDB database a document oriented NoSQL database that does not require a schema definition before usage All the artefacts are stored in the database except for files which are stored and managed by the underlying file system Below is the schema of the database used by the web service Figure 3 4 4 TIS Internet Information Services a set of Internet based services for servers using Microsoft Windows 4 MongoDB http www mongodb org 28 A Website for A
62. ng frame ORF analysis to correlate which open reading frames are more present in tissue samples with certain diseases and try to determine if those ORFs have a greater impact in causing that disease 4 1 1 Obtaining the Junctions Type Count In order to get the number of gene types from a SAM BAM file one needs an annotation file of the whole human genome GTF An up to date version of the human genome sequence can be obtained from the GENCODE project web page The GTF format has a field named gene_status that indicates if the gene is known or new These fields are present in both the annotation file GIF and the mapped read file to analyse BAM 6 The GENCODE project http www gencodegenes org 37 Alternative Splicing Analysis Process e The name of the chromosome e The starting position of the feature in the chromosome e The ending position of the feature in the chromosome In order to count the number of each type of function we must count the gene status field which can be one of these values e KNOWN the gene is known e NOVEL the gene is new e PUTATIVE the gene is not known but it is believed to be a gene by its open reading frame To achieve this we run the following command to convert a BAM file to a BED with12 fields samtools view h infile bed awk if 6 N 1 SQ print samtools view bS bamToBed bed12 i stdin gt outfile bed This command will rearrange the dat
63. nously Jobs are defined by a name and a command to execute with a list of parameters whose meaning varies depending on the job For a download job the parameters are the download URL and an optional new file name Currently there are these types of job e Bam2Sam Converts a BAM file to the SAM format as seen in Figure 3 6 e Download Downloads a file to a folder from an URL e InterProScan Uses the InterProScan sequence search web service to scan FASTA files for known proteins e Orf Find Runs an ORF finder job using a FASTA file as input and outputs an ORF file e Sam2Bam Converts a SAM file to the BAM format e Sam2Fasta Converts a SAM file to the FASTA format Jobs Add Convert Bam to Sam Job Add Cancel Figure 3 6 BAM to SAM conversion job example 47 InterProScan web service http www ebi ac uk Tools webservices services pfa iprscan5_soap 31 A Website for Alternative Splicing Analysis The worker bot checks the command name and constructs the command to execute and takes care of downloading the necessary files from the server and uploading the newly created files to the server A job can have an associated project in which case the projects associated files can be used as input and the output is saved in the project s folder Because of the generic nature of the job not having specific information such as the command line to execute but the action to perform and the list of param
64. ocalhost 8080 e MongoDBHost Name of the MongoDB server host Default is localhost e MongoDBPort Name of the MongoDB server port Default is 27017 e MongoDatabase Name of the MongoDB database name Default is WebServiceDB e PageSize This applies to all views that list items and defines the number of items to display per page Default is 10 e DataFolder Name of the folder to store the data files Default is data e StdioFolder Name of the folder to store the command s stdio files Default is stdio These settings can be configured in the WebService Self exe config file Figure A 1 lt xml version 1 0 encoding utf 8 gt lt configuration gt lt appSettings gt lt add key DataFolder value data gt lt add key StdioFolder value stdio gt lt appSettings gt lt configuration gt Figure A 1 Web service example configuration file The data and stdio folders must exist on the disk and have read and write permission from the operating system It is recommended that these folders be symbolic links if the 72 NET framework application configuration files http msdn microsoft com en us library ms229689 80 Symbolic links http en wikipedia org wiki Symbolic_links 49 Web Service User Manual folders are on a different disk or a network drive to avoid problems that might arise from parsing unusual paths For security reasons the web service doesn t allow the c
65. of this thesis as we will focus on the computational platform that will facilitate the execution of tasks related to this process Our work starts after the genome assembly but uses information about the reads distribution Image taken from Wikipedia http en wikipedia org wiki DNA_ Translation Basic Concepts and Survey on Technology A gene as seen in Figure 2 2 is composed of a sequence of smaller parts These parts are of two types exons and introns Exons and introns alternate in sequence to make up the complete gene Only the exons convey information to encode the proteins as introns are noncoding regions It happens that at different times and circumstances a different set of exons may be active may be translated This means that the same gene may produce different products This phenomenon is called alternative splicing Some of the alternative products may cause diseases Exon 1 Intron 1 Exon 2 Intron 2 Exon 3 Intron 3 Exon 4 IAMTHEQSRTVFIRSTQPBASTRVSECONDQABDECTVSENTENCE AM THE FIRST SENTENCE AM THE SECOND SENTENCE Figure 2 2 Principles of alternative splicing This process happens inside our cells and is responsible for our physical traits and some diseases such as certain types of cancer To assess which genes are active in each organism one must take the initial assembled genome or parts of it and count the number of times a certain sequence appears in it The more times a sequence appears the more ac
66. olutions named in Chapter 2 were studied before starting the implementation but they were dismissed because they were not quite what was required The NET framework was chosen because it has many of its components available as open source with a big user community and good development tools The Nancy framework was chosen because of its simplicity and ease of use and for allowing for a bit more freedom in comparison to standard model view controller MVC frameworks Other technologies and frameworks were considered namely node js and other free frameworks but were dismissed either because of their poorer development tools or because of difficulties achieving true cross platform targeting because of OS specific modules in these frameworks The same binary file is used to run on both Windows and Linux machines without modifications MongoDB is used as a database because the application does not require the full power of a SQL database and because it is simpler to model the database as we do not need to explicitly define a data schema It also allows us to store variable arrays more easily in comparison to a relational database as it is the case with a job s parameter list File management uses the file system for storing files This allows us to use the underlying operating system s file API to manage files The worker bot transfers files from and to the web service s server This adds overhead to the process but allows for more flexibili
67. orkspace for each user EMBOSS EMBOSS is The European Molecular Biology Open Software Suite EMBOSS is a free Open Source software analysis package specially developed for the needs of the molecular biology e g EMBnet user community The software automatically copes with data in a variety of formats and even allows transparent retrieval of sequence data from the web Also as extensive libraries are provided with the package it is a platform to allow other scientists to develop and release software in true open source spirit EMBOSS also integrates a range of currently available packages and tools for sequence analysis into a seamless whole EMBOSS breaks the historical trend towards commercial software packages Because EMBOSS is a collection of tools for the most common bioinformatics operations there are a few web interfaces that use it and provide user account management and a workspace per user There are also some workflows or pipelines that use EMBOSS and work similar to job schedulers as they allow job management and distributed computing Other There are many other job scheduling solutions that offer many features common to these types of problems like e Script storage where missing scripts are automatically transferred to the target system e Event driven where jobs are ran when a worker bot is available e Agents where host systems install a program that connects to the central job server e Multi platform e
68. pplications involving the analysis of genomic data require powerful computational resources In this chapter an architecture and implementation of a computational platform that we think is adequate for genomic scale data analysis is presented We first describe the proposed architecture then we present the functionalities and lastly the implementation technologies and choices made 3 1 Web Service In bioinformatics the tasks described in the previous chapter are not user friendly and are error prone if the person executing the task is not familiarized with the tools involved In order to streamline the process of executing these tasks and other tasks commonly used in bioinformatics a web service was developed to tackle the most common usage scenarios and user s needs 3 1 1 Main Use Cases Before starting the development of the web service a set of requirements both functional and non functional and use cases Figure 3 1 were made regarding the web service namely e The solution must be cross platform In order to run from any computer the proposed solution was to build a web service since it will be accessible from any system with a web browser 24 A Website for Alternative Splicing Analysis e The solution must be user friendly Since the target audience is not tech savvy the solution should be simple to use and hide the inherent complexity in the underlying process e The solution must work well with tablets
69. reateDate Timestamp with the creation date e IsAdmin Boolean value indicating whether the user is an administrator or not e IsCluster Boolean value indicating whether the user is a cluster account or not e Pending Optional boolean value indicating whether the user account is pending validation by an administrator account MongoDB allows the creation of arrays as values of a field This is used in the job s args field to store a variable length array of strings that are used as arguments for the job s command 3 2 2 Web Service Architecture The web service requires an account to be able to create new jobs Most operations require authentication before being used A menu bar is shown at the top of the page with access to the various functionalities as seen in Figure 3 5 Web service Welcome admin w profile My jobs My projects Welcome to the web service Logout Figure 3 5 Web service initial menu 30 A Website for Alternative Splicing Analysis A user can create an account from the web service An admin account must approve new user accounts before these can be used for security reasons Administrator accounts must be created manually for security reasons by a person with direct access to the database Cluster accounts are a special kind of account used by worker bots to ping the service for pending jobs We can have more than one worker bot account configured to execute more jobs asynchro
70. reation of administrator accounts Users cannot create new user accounts without an administrator first allowing such an account Administrator accounts must be created manually by directly importing users to the database This can be done using mongoimport a program that comes with MongoDB mongoimport host localhost 27017 d WebServiceDB collection users file users json upsert This command will import users to the default database WebServiceDB on the default host localhost into the users collection from a JSON file users json The users json file looks like this id oid 000000000000000000000001 login admin name Administrator passhash 753068535f 964205070a59af8a0c64aacc9883d03febd7ab8d2b92ed29c3dd93 createdate date 2014 01 01T00 00 00 000 0100 isadmin true Please note that the oid must not be 0 otherwise the system will stop working correctly Notice the isadmin boolean set to true indicating an administrator account The passhash provided is for the password demodemo You can change the password using the web service The password hash stored in the database is composed of a SHA256 hash of the SHA1 password hash with a salt appended to it The password salt value is defined in WebService Common Settings cs as PassSalt PassHash SHA256 SHAI password PassSalt The web service uses Twitter Bootstrap theming version 3 for its user interface The d
71. regulation in RNA Seq samples It accepts aligned RNA Seq reads and assembles the alignments into a parsimonious set of transcripts Cufflinks then estimates the relative abundances of these transcripts based on how many reads support each one taking into account biases in library preparation protocols TopHat TopHat Trapnell Pachter and Salzberg 2009 Langmead Trapnell et al 2009 Kim and Salzberg 2011 Kim et al 2013 is a fast splice junction mapper for RNA Seq reads It aligns RNA Seq reads to mammalian sized genomes using the ultra high throughput short read aligner Bowtie and then analyses the mapping results to identify splice junctions between exons Bowtie Bowtie Langmead Trapnell et al 2009 Langmead Schatz et al 2009 Trapnell Pachter and Salzberg 2009 is an ultra fast memory efficient short read aligner It aligns short DNA sequences reads to the human genome at a rate of over 25 million 35 bp reads per hour Bowtie indexes the genome with a Burrows Wheeler index to keep its memory footprint small typically about 2 2 GB for the human genome 2 9 GB for paired end A more recent version named Bowtie 2 Langmead and Salzberg 2012 Langmead Trapnell et al 2009 exists 14 Cufflinks http cufflinks cbcb umd edu 15 TopHat http tophat cbcb umd edu 16 Bowtie http bowtie bio sourceforge net 17 Bowtie 2 http bowtie bio sourceforge net bowtie2 13 Basic Concep
72. rform more operations These standards are referred to as web APIs application programming interfaces Many standards exist that sit on top of these technologies They vary in the way that the web service s API is specified and used and in which format the messages should be transferred The two most widely used formats to transmit structured documents or data objects over the internet are JSON and XML Another big advantage of having web services sit on top of HTTP is that the same API can be used by both web browsers and programs that consume that API If the web service detects that a web browser is making the request it can return an HTML document to display in the browser If the web service detects that a consumer application is making the request it can return a simpler response with the data in one of the named document formats to be processed by the application in its own way Detecting which client is accessing the API can be done either by analysing the HTTP user agent a header in the protocol that identifies the browser or HTTP library being used or by having the programs that consumes that API specify that they do not want a visual response in HTML but a JSON or XML response to process in their own way This can be done in a number of ways and varies depending on the standard used 3 F measure http en wikipedia org wiki F measure 22 HTTP protocol http en wikipedia org wiki HTTP 33 JSON http en wikipedia org w
73. sses requiring the execution of several tasks a software tool capable of collaborative work among researchers sharing any information they think useful to others a ubiquitous computational tool accessible from any place with access to the internet The work was tested in a very specific domain namely with procedures 40 Conclusions mainly used to test the impact of aberrant alternative splicing in cancer The proposed and developed platform enables running standard tools used in aberrant alternative splicing studies the possibility of chaining tools to build a pipeline that accepts the aligned reads as input and produces the aberrant alternative splicing results an automatic process using the API from the adequate web resource to fetch the final information of the analysis concerning the domains of missing parts of the proteins encoded by the gene under analysis using the InterProScan web service 5 2 Future Work Few issues were left unresolved in the web service Most features were implemented There are however some aspects that can be improved namely the web service could return different views depending on the command executed and show the data in a more obvious way for the user to analyse Although this was not implemented it is fairly easy to do and merely requires the job view route in the web service to return a different view depending on the command name and create the custom views using HTML and JavaScript Another i
74. ssue that exists is the download command that is executed remotely only to have the downloaded file sent back to the server This is because of the distributed nature of the web service and the fact that the download command is a job An option to run downloads locally could be added in the future When transferring files from the web service worker bots will overwrite existing files with the same name in their workspace even if these are the same as the ones in the server An improvement could be done by checking the file sizes and modification dates or by using an error detection code or a cryptographic hashing algorithm to verify that the files are the same Regarding the existing types of jobs more jobs useful for alternative splicing analysis could be added to add more functionality to the solution Adding unit testing to the web service s public API and to helper functions is a future improvement to ensure it works correctly with future changes The web framework used provides a way to test the web service and one of the many unit testing libraries can be used to test other code Although the complete set of modules of the pipeline starting at the raw RNA Seq set of unaligned reads and ending at the identification of the missing domains of proteins is very easily accommodated in the proposed framework we have concentrated only on the pipeline stages concerning the analysis of the gene expression generation and analysis of ORFs To ma
75. technology molecular biology has seen an incredible increase in the number and diversity of software developed to solve a wide range of biological problems and to speed up a large number a tedious task expert biologist have to perform Data analysis in molecular biology has also seen considerable advancements with the crucial help of informatics Despite the reported advancements in software applied to biological issues there are important problems that have to be addressed For complex studies the required information is scattered among a lot of different sources in the internet and encoded in different Introduction formats Complex analysis using genetic data for example requires powerful computational resources The available software used in complex analysis may encompass the use of a lot of tools the user must know how to use them all There are a lot of complex analyses that have to be done routinely by biologist To all of these problems bio informatics may give a very useful contribution as we expect to show with our work 1 2 Motivation and Goals Advances in the study of the effect of aberrant alternative splicing in cancer may have an enormous social impact Insights on the mechanism that originates cancer may suggest processes to prevent or reduce its occurrence There is also an enormous amount of both data and software tools available on the web The software available is usually collections of programs that solve a small part
76. tform Chapter 5 concludes this thesis and outlines future improvements Chapter 2 2 Basic Concepts and Survey on Technology We first introduce basic biological concepts necessary to understand the scope of this thesis We then describe the process of sequencing programs and the state of the art in algorithms used to study the impact of alternative splicing The rest of the chapter addresses the technological tools and alternatives to implement the computational platform 2 1 Basic Biological Concepts The genome is composed by a set of genes and it contains all of our hereditary information The genes are found in chromosomes and made of deoxyribonucleic acid DNA Genes determine the various characteristics of all living organisms by telling our cells how to make proteins as seen in Figure 2 1 Proteins do not give a human being big ears per se but their production is determined by your genes and that will be responsible for one s physical traits Basic Concepts and Survey on Technology newly born protei amino acids large subunit Figure 2 1 Diagram showing the translation of mRNA and the synthesis of proteins by a E 1 ribosome Genes are made of DNA and the sum of an organism s DNA is the genome DNA is made from different combinations of four base molecules ACGT repeated over and over again in different configurations It is nowadays technically possible to determine the genome of an individual To do th
77. tive that particular gene is Once that count has been established one can use it to compare to known traits and assay the probability of developing certain diseases Software tools help us produce these counts from reads and others help us view and analyse the necessary information in order to produce a report about the functioning of those particular genes The output of the RNA sequence one of the techniques for doing genome sequencing is a histogram of the reads which tell us which genes or parts of the genes are more active or inactive This is useful information to study genomic based diseases 2 2 Software for the Biological Analysis Software for the Biological analysis is divided by function Most programs abide by the UNIX philosophy where each program is designed to do one thing only and do it well Because of this most programs are command line programs that take the input of others and produce a new set of outputs for another program There are many available tools each dealing with one or more steps of the process notably Basic Concepts and Survey on Technology e Read aligners align reads to a reference genome such as the human genome e Read mappers work similarly to read aligners but try to identify splice junctions e Differential expression programs assemble reads and estimate their abundance for further analysis e Visualizers take input from differential expression analysis and present commonly used visual
78. to use POST for both adding elements and editing them Distinction between both operations is then done by appending edit to the end of the URL This is not a specification but a common usage style These modifications have many reasons to be namely firewall restrictions or web browser incompatibilities 35 Representational state transfer http en wikipedia org wiki Representational_state_transfer 20 Basic Concepts and Survey on Technology 2 5 2 Web Frameworks Many web frameworks support writing RESTful web APls Because many of the described steps are common and repetitive they abstract away the burden of implementing REST from scratch using only an HTTP framework and provide much functionality that eases the development of these APIs Some of these common functionalities are e Routing Allows the specification of the URL and which method should be executed for each request It takes care of figuring out which route will handle which request e Model binding Allows sending of data in an easy way as a response to a request e View engines Allows the usage of view engines that make it easier to create dynamic web pages e Localization Makes it easy to have web pages translated into multiple languages without having to modify a lot of files in the process e Testing Provides mechanisms to test REST routes to make sure they are working properly e Content negotiation Allows the client to negotiate what type of cont
79. ts and Survey on Technology Other There are many other software packages for the various stages of the process that won t be enumerated here because of their large number including variations of these made to scale in computer clusters On the SEQAnswers wiki an on line community for next generation sequencing there are more than 600 software packages listed for solving one or more stages of the processes involved in next generation sequencing 2 2 3 Databases The ENCODE project aims to identify all functional elements in the human genome sequence There are several types of data referenced including RNA Seq data We can obtain data aligned to areference genome from there There are various cell lines referenced in the ENCODE project that are taken from human tissue samples and later transformed in a laboratory so that they can be replicated and researched throughout various laboratories around the world Some of these cell lines are derived from cancer tissues You can find more information about the different cell lines on the project s page This is the description from the two samples we will use taken from the project s page K562 and GM12878 K562 is an immortalized cell line produced from a female patient with chronic myelogenous leukemia CML It is a widely used model for cell biology biochemistry and erythropoiesis GM12878 is a lymphoblastoid cell line produced from the blood of a female
80. ty as web can have the server running on a standard computer but the worker bot running on a faster machine Because worker bots have their own accounts we can have multiple worker bots executing pending jobs in parallel by having several worker bots configured and thus have a more scalable solution Note however that due to the transfer of files from and to the central server this one might become a performance bottleneck due to the network traffic involved in the file transfers and the internal sharing of the same hard disk drive It is therefore recommended not to use a high number of worker bots on the same web server 55 Mono http www mono project com 56 Job scheduler http en wikipedia org wiki Job_scheduler 37 NET foundation http www dotnetfoundation org 3 Node js software platform http nodejs org 34 Chapter 4 4 Alternative Splicing Analysis Process This chapter describes the application of the developed computational platform to some of the steps taken to analyse data for the problem of aberrant alternative splicing in cancer The data as mentioned in Chapter 2 was downloaded from the ENCODE project that aims at the identification of all functional elements in the human genome The project hosts various cell lines taken from human tissue samples and later transformed in a laboratory to be replicated and researched throughout various laboratories around the world The samples we use are cancer
81. usar por parte dos especialistas das ci ncias vivas A plataforma tamb m muito f cil de gerir ao permitir utiliza o atualiza o remo o de recursos em diferentes sistemas operativos e sem impacto no processo de an lise Tamb m permite a execu o individual de cada passo da sequ ncia de tarefas Apesar de a plataforma ser de uso geral a vers o atual est otimizada para a aplica o de m todos que usam informa o de alto rendimento de transcriptomas de amostras de cancro para identificar ocorr ncias de splicing alternativo e o seu impacto na estrutura transcrita A plataforma desenvolvida foi testada em casos de estudo usando diferentes linhas celulares de bases de dados dispon veis na p gina web do projeto ENCODE iii iv Acknowledgements I would like to thank my supervisor at FEUP Professor Rui Carlos Camacho de Sousa Ferreira da Silva for his suggestions and reviews on both the solution developed and this document I would also like to thank my co supervisor from the Swiss Institute of Bioinformatics Doctor Pedro Gabriel Ferreira for his help in understanding the pipelines related with alternative splicing analysis Vitor Moreira vi Contents 1 Introduction ssssssssssssssssssesssennennennennensnnsnsnsessnssesesesseensse 1 1 1 CONTEXT O sn ester untar an ni as Permanent 1 1 2 Motivation and Goals cccccesccesscessceesceeeceeeeeeceeeseeeseeeseeeseeenaeeaees 2 1 3 PORC A tree ner a
82. w user account A 3 1 Sign In You can sign in using the sign in link in the top right corner of the home page You need to enter the following fields to sign in Figure A 3 e Login Unique account login Must be at least 3 characters long e Name User name Must be at least 3 characters long e Password Password used to authenticate into the web service Must be at least 4 characters long Users Create Account Add Cancel Figure A 3 Sign in form 52 Web Service User Manual A 3 2 Login In order to access the web service you must first login into the web service You can do this by using the top menu login and entering your login and password Figure A 4 If you try to access a restricted page you will be redirected to a sign in page Login Figure A 4 User account menu One logged in you will see a welcome message with access to your user profile where you can change your password your jobs and projects and a logout form Figure A 5 Welcome admin View profile My jobs My projects Logout Figure A 5 Login menu 53 Web Service User Manual A 3 3 Web Service Menu After logging into the web service you have access to a menu with all the functionality provided by the web service Figure A 6 Web service Figure A 6 Web service menu The menu gives access to the following items e Clusters Manages cluster accounts used by worker bots e Files

Download Pdf Manuals

Related Search

Related Contents

traser  APÊNDICE A – FORMULÁRIO DE CONSENTIMENTO  Elettroterapia  The micrOMEGAs user`s manual, version 4.0 - LAPTh  Maytag PSH4BG Technical Literature    WiReLeSS ViDeO HOMe MONiTOR Getting Started  Bedienungsanleitung  アイスツール取扱説明書 株式会社ロストアロー  Acctiva Easy 1202, 1204, 1206, 2403 Acctiva Easy 6/12  

Copyright © All rights reserved. Failed to retrieve file