User's guide to ChIP-Seq applications in the Vital
Contents
1. 1 o merged sga gt CTCF peaks out The public data which are accessible via the ChIP Seq server menu are stored in the so called MGA Mass Genome Annotation data repository located in db mga All data in the MGA repository are provided in SGA format At the first hierarchical level the mga root directory is split into sub directories corresponding to genome assemblies e g hg 18 mm9 at the second level according to the data series e g hg18 barski07 A third level is used for ENCODE data sets e g hg18 encode GSE25416 A data series sub directory typically contains data from one publication and or one series in GEO GSE entry 3 Auxiliary Tools for data reformatting We also provide a series of auxiliary tools most of them perl scripts that can be used to perform format conversion tasks These perl scripts are installed in mnt common perl However some of the third party reformatting tools are not installed on all Vital IT machines We therefore recommend that you carry out all reformatting tasks on the ChIP Seq server machines directly Most of the ChIP seq data sets come in BAM or BED formats If you have BAM files and you need to convert them to SGA format the steps to follow are 1 Log into the ChIP Seq auxiliary server machine on Vital IT ssh l username ccg serv02 vital it ch 2 Make sure that the directories home local bin and home local perl are in your PATH environment variable e g on a bash shell com2. CF out where CTCF sga is the input file containing the list of mapped CTCF tags A CTCF is the reference feature CTCF plus strand B CTCF is the target feature b is beginning of the range considered in the output histogram e is the end of the range W is the window width C is the count cut off value n is the normalization mode 1 means count density CTCF out is the output file containing the histogram values in text format The input file for this example can be found at db mga hg18 barski07 CTCF sga Documentation for ChIP Seq programs is provided via UNIX style man page type e g man chipcor Short usage instructions can also be obtained by simply typing the name of the program without any options or arguments In the above example the reference and target features were contained in the same input file This is not always the case In the following example we will analyze the distribution of CTCF peaks in mouse embryonic stem cells relative to transcription start sites from the Eukaryotic Promoter Database EPD The input data for this analysis are provided in two separate files which can be found at db mga mm9 chen08 ES_CTCF_peaks sga db mga mm9 epd Mm_EPDnew_001_mm9 sga These files need first to be merged before they can be processed by chipcor sort s m k1 1 k3 3n k4 4 Mm_EPDnew_001_mm9 sga ES_CTCF_peaks sga gt merged sga chipcor A TSS B CTCF _P b 2500 e 2500 w 50 c 1 n
3. ED file If the strand field which is optional in BED is presented the conversion will be done in oriented mode Otherwise the conversion will be done in centered mode and the strand field will be set to zero However centered mode can be forced by the command the line option c The most common command pipeline to perform BED to SGA conversion is the following bed2sga pl f lt feature gt s lt species gt reads bed sort s k1 1 k3 3n k4 4 compactsga gt reads sga There are several additional reformatting programs that may be useful in certain situation The C program featreplace replaces all feature names in an SGA file by a new name featreplace f lt feature gt old sga gt new sga The liftOver program from UCSC can be used to convert chromosomal coordinates in a BED file from one genome assembly to another e g liftOver input bed db liftOver nm amp ToMm9 over chain gz output bed trash Note that liftOver chain files can be found in db liftOver Additional conversion scripts include sga2bed pl SGA to BED sga2wigFS pl SGA to WIG wiggle fixedStep format sga2wigVS pl SGA to WIG variableStep format The latter two programs produce custom track files that can be uploaded to the UCSC genome browser for visualization 4 Working on the Web with your own server resident data If you would like to use the ChIP Seq server web interface http ccg vital it ch chipseq with your own large data files we strongly r
4. User s guide to ChIP Seq applications in the Vital IT environment 1 Basics about the ChIP Seq Tools The ChIP Seq software provides a set of tools performing common genome wide ChIP Seq analysis tasks including positional correlation analysis peak detection and genome partitioning into signal rich and signal poor regions These tools exist as stand alone C programs and perform the following tasks I Positional correlation analysis chipcor Il Tag centering chipcenter II Signal peaks detection chippeak IV Identification of signal enriched regions chippart V Feature extraction tool chipscore Because the ChIP Seq tools are primarily optimized for speed they use their own compact format for ChIP Seq data representation called SGA Simplified Genome Annotation SGA is a line oriented tab delimited plain text format with the following five obligatory fields per line I Sequence ID char string II Feature char string IH Sequence Position integer IV Strand or 0 V Tag Counts integer An additional field may be added containing application specific information used by other programs In the case of ChIP Seq data an SGA file represents the genome wide tag count distributions from one or several experiments However the format can also represent a large variety of other types of genomic data including derived features such as peaks extracted from ChIP Seq data or genome annotations such as promoters An
5. e position and strand For instance the two input lines NC_000001 9 H3K4me3 5011 1 NC_000001 9 H3K4me3 5011 1 would be replaced by the following single line NC_000001 9 H3K4me3 5011 2 Compacting sga files saves space and reduces program execution time but unlike sorting is not formally required by the ChIP Seq tools The bed2sga pl has two basic modes of operations centered and oriented In the centered mode the midpoint between the start and end position from the BED line 2 and 3 field will be used as position In the oriented mode the conversion depends on the strand indicated in the BED file 6 field If the strand is then the value of the start field will be incremented by one and used as position in the SGA file If the strand is the value of the end field will be used as position in the SGA file Incrementing the start position in oriented mode is necessary because the BED format has a zero based numbering system for chromosomal regions whereas SGA has a 1 based numbering system The behavior of the two conversion modes is illustrated by the following example BED input chr1 100000266 100000291 chr1 100000383 100000408 0 O SGA output centered mode NC_ 000001 9 CHIPSEQ 100000278 1 NC_000001 9 CHIPSEQ 100000395 1 SGA output oriented mode NC_000001 9 CHIPSEQ 100000267 1 NC_000001 9 CHIPSEQ 100000408 1 By the default the conversion mode depends on the contents of the B
6. ecommend that you store your data directly on the web server machine You can then access your file via the input menu option Server resident SGA Files by Filename By choosing this data access mode you can avoid transfer of large files over the network To be able to store your data directly on the server proceed as follows 1 Send an email request to ccg mail isrec isb sib ch indicating your Vital It username We will then create a repository for your data on the server machine ccg serv01 vital it ch in data scratch username to which you will have read write access Note that files in this area will be permanently stored but not backed up 2 Create a sub directory corresponding to the genome assembly of your data The name of the subdirectory should conform to the UCSC nomenclature e g hg18 hg19 mm9 dm3 ce6 sacCer3 3 Transfer your data to the appropriate subdirectory on ccg serv01 vital it ch Note that only files in SGA format can be read from this location 4 On the Web interface select your file via the menu Server resident SGA Files by Filename by typing username assembly myreads sga e g gambrosi hg18 All_histones sga into the text area
7. mand line type PATH PATH home local bin home local perl export PATH 3 Find out which genome assembly has been used to generate the BAM files and make sure that the chromosome names agree with the naming scheme of the UCSC genome browser Then use the following type of command bamToBed i reads bam bed2sga pl f lt feature gt s lt species gt sort s k1 1 k3 3n k4 4 compactsga gt reads sga The program bamToBed belongs to the BEDTools package a suite of utilities for comparing genomic features that can be installed from http code google com p bedtools bamToBed converts BAM alignments to BED format The bed2sga pl tool is a perl script that converts BED files to SGA format To find instructions how to use it type bed2sga pl h SGA format requires a feature field the second field containing a character string that defines the genomic feature represented by the file The BED file does not have an equivalent field You therefore need to supply a feature for the conversion via the f option The species is the name used by UCSC for the genome assembly to which the reads have been mapped e g hg18 mm9 dm3 etc bed2sga pl will automatically convert UCSC chromosome names into corresponding NCBI RefSeq accession numbers The reasons why sorting is required have been explained before The program compactsga is a C program which merge lines corresponding to the same genome position identical sequence nam
8. stance constitute an example of an un oriented feature Un oriented features are identified by a 0 zero in field 4 The counts field contains the number of reads that have been mapped to a specific base position on the chromosome An example of an SGA formatted file is shown here below NC_000001 9 H3K4me3 4794 1 NC_000001 9 H3K4me3 6090 1 NC_000001 9 H3K4me3 6099 1 NC_000001 9 H3K4me3 6655 1 NC_000001 9 H3K4me3 18453 1 NC_000001 9 H3K4me3 19285 1 NC_000001 9 H3K4me3 44529 1 NC_000001 9 H3K4me3 46333 1 NC_000001 9 H3K4me3 46349 1 Chip Seq programs require SGA files to be sorted by sequence name position and strand In a UNIX environment the command to properly sort SGA files is the following sort s kl 1 k3 3n k4 4 lt SGA file gt 2 Using ChIP Seq Tools on the Vital IT platforms To use the ChIP Seq tools from the command line login to a Vital IT machine e g ssh l gambrosi frt el vital it ch The ChIP Seq programs are installed on Vital IT in the directory mnt local bin Make sure that this directory is in your path To add it to your PATH environment variable from a bash shell type PATH PATH mnt local bin export PATH You invoke the programs by simply typing the program name i e chipcor chipcenter chippeak chippart chipscore Here is an example of using the ChIP Seq correlation program chipcor chipcor A CTCF B CTCF b 1000 e 1000 w 1 c 1 n 1 CTCF sga gt CT
9. y type of genomic feature that can be projected to a single base on a chromosome can be represented in SGA format The sequence field typically identifies a chromosome The software does not impose any naming convention However if you merge data from different experiments the same names must be used for the same chromosomes The public data files access by the ChIP Seq server menu use NCBI RefSeq accession numbers as sequence chromosome identifiers Chromosome names as used by the UCSC genome browser constitute another de facto standard and are readily converted into NCBI RefSeq accession numbers However keep in mind that chromosome names ambiguous unless the corresponding assembly is indicated Do not mix chromosome names from different assemblies otherwise you will get wrong results The feature field contains a short code which identifies an experiment It often corresponds to the name of the molecular target of a ChIP Seq experiment Its function is to distinguish data lines relating to different experiments that were merged into a single file The position field contains the position within the sequence The strand field indicates the strand to which the feature has been mapped The SGA format distinguishes between oriented features that occur either on the plus or on the minus strand of the chromosome sequence and un oriented features which cannot be assigned to one or the other strand Peaks from a ChIP Seq experiment for in
Download Pdf Manuals
Related Search