The UEA sRNA Toolkit: A User Guide for the Perl Implementation
Contents
1. 43 3 10 Plant Target Prediction Tool This tool identifies sRNA targeted transcripts The rules used for tar get prediction are based on those suggested in i and 25 Specifically miRNA target duplexes must obey the following rules e No more than four mismatches between sRNA and target G U bases count as 0 5 mismatches e No more than two adjacent mismatches in the miRNA target duplex e No adjacent mismatches in positions 2 12 of the miRNA target duplex the positions are indexed starting with the 5 end of the miRNA e No mismatches in positions 10 11 of miRNA target duplex e No more than 2 5 mismatches in positions 1 12 of the of the miRNA target duplex e MF EmiRNA target 2 0 74 MFEmiRNA miRN Ax Parameters e Required transcriptome The location of the transcriptome file in FASTA format out The path of the output directory Either pasted srnas or srna file containing sequences in FASTA format pasted_srnas should contain no more than 50 se quences If srna_file is provided it should specify the loca tion of the sRNA file Allowed nucleotide symbols A G C T U N Example srna tools pl tool target out output t pasted_srnas gt a GCTTCTATCTTTTTCTTTCGTGCT transcriptome arabidopsis fa A target prediction results file looks as shown below gt AT4G33780 1 287 309 Symbols FUNCTIONS IN molecular_function unknown INVOLVED IN biological_pr2. This can be also due to the file missing in a new version of miRBase in which case users should contact miRBase for help Can not move results to output directory Ensure that the output directory exists prior to running the toolkit Error reading input Fasta file Some tools e g miRCat require the sRNA sequences in redundant format Please check the specifications of each tool prior to running it Error with Parameters All parameter names must be precedeed with a Using a single dash will not work as it is interpreted as a negative value 48 4 2 Tool Specific Errors 4 2 1 Sequence File Pre Processing Tool How many sequences are retained and how many are removed A summary text file is created when the tool is run containing details of how many sequences matched the adaptor and how many were not within the size limits If a large number of sequences do no match the adaptor it is advisable to check the input file on the command line using the Unix command less Also make sure that the adapter sequence is spelt correctly and that only a prefix for the 3 adapter or a suffix for the 5 adapter is given 4 2 2 Filter Tool Even with no filter options some sequences are removed Low complexity sequences which by their nature are likely to match to the genome many times are removed by default to avoid an unnecessarily large PatMaN file being created A sequence of interest was removed by the t rRNA filter Th
3. Example srna tools pl tool mircat genome data arabidopsis fa srna_file files GSM118373_Rajagopalan_leaf fa out output m genomehits 5 hit_dist 100 maxgaps 4 max_overlap_length 50 max_percent_unpaired 60 max_unique_hits 4 maxsize 24 min_abundance 6 min_energy 10 0 min_gc 20 min_hairpin_len 80 min_paired 25 minsize 19 no_complex_loops percent_orientation 80 pval 0 2 trrna window_length 100 miRCat returns the results as a zip file containing the following output files e A csv file showing all predicted miRNA loci it displays the following information about each predicted miRNA Chromosome Start position End position Strand orientation Abundance number of times sequenced in high throughput dataset Sequence of predicted mature miRNA Representative sequence accession from input dataset Length of predicted mature miRNA Number of matches to genome Length of predicted precursor hairpin sequence G C content of hairpin sequence Minimum free energy MFE of predicted hairpin sequence Adjusted MFE MFE MFE lengthhairpin 100 Shows MFE per 100nt making results comparable randfold p value miRNA shows predicted miRNA sequence s if any along with abundance in input dataset shown in brackets e A text file containing predicted miRNA precursor sequences and struc tures in dot bracket notation e A pdf
4. counts for sRNAs matching miR156 exactly and with 1 3 mismatches are combined into one group organisms If defined matches to the same miRNA in dif ferent organisms are combined into one group_variant If defined matches to different variants of the miRNA are combined into one such as 25 x different mature sequences that can arise from the same pre cursor annotated in miRBASE as 3p 5p s or as in the ID of the miRNA and applies to mature sequences only x different precursors that produce the same mature sequence annotated as 1 2 etc in miRBASE See miRBASE help for more details keep_best If defined only the best matches are kept for each sRNA sequence For example if there are miRNAs with a per fect match for a s RNA no miRNAs from the same organism with any mismatches would be accepted for the same sRNA This is not applied to miRNA matches from different organisms Often sRNAs will match multiple members of the same miRNA family This option helps to reduce the complexity of the output for those cases maxsize The maximum sRNA length 18 lt mazsize lt 35 default maxsize 25 minsize The minimum sRNA length 18 lt minsize lt 35 default minsize 18 mismatches The maximum allowed number of mismatches 0 lt mismatches lt 3 default mismatches 0 overhangs If defined mirProf will accept overhanging 5 or 3 bases as mismatches providing the mirbase database has bee
5. must be at least 17 of the 25 nucleotides centered around the miRNA The hairpin must be at least 75nt for plants or 50nt for ani mals in length The percentage of paired bases in the hairpin must be at least 50 of base pairs in the hairpin this threshold can be adjusted using the max_percent_unpaired parameter e The hairpin with the lowest minimum free energy MF E from the se quence windows is then chosen as the precursor miRNA pre miRNA candidate e The pre miRNA candidate is then tested using randfold using a de fault cutoff of 0 1 this threshold can be adjusted using the pval parameter miR164 poy Ou jj oros roco909 x bad TILS EL gt SDF MS ESPOSO e A miR164 Figure 3 3 RNA fold output showing miR164 precursor Parameters e Required genome The location of the genome file in FASTA format srna_file The location of the sRNA file in FASTA format out The path to the output directory e Optional genomehits The maximum number of genome hits 1 lt genomehits default genomehits 16 20 hit_dist The maximum distance between consecutive hits on the genome 0 lt hit_dist default hit_dist 200 max_gaps The maximum number of consecutive unpaired bases in miRNA region 0 lt max_gaps lt 5 default max_gaps 3 max_overlap_length The maximum total length nt of overlap ping sRNAs 30 lt maz_overlap_length default 70 max_perc
6. Schmid and Detlef Weigel Specific effects of micrornas on the plant transcriptome Dev Cell 8 4 517 527 Apr 2005 Frank Schwach Simon Moxon Vincent Moulton and Tamas Dalmay Deciphering the diversity of small rnas in plants the long and short of it Brief Funct Genomic Proteomic 8 6 472 481 Nov 2009 David Swarbreck Christopher Wilks Philippe Lamesch Tanya Z Be rardini Margarita Garcia Hernandez Hartmut Foerster Donghui Li Tom Meyer Robert Muller Larry Ploetz Amie Radenbaugh Shanker Singh Vanessa Swing Christophe Tissier Peifen Zhang and Eva Huala The arabidopsis information resource tair gene structure and function annotation Nucleic Acids Res 36 Database issue D1009 D1014 Jan 2008 P M Waterhouse M B Wang and T Lough Gene silencing as an adaptive defence against viruses Nature 411 6839 834 842 Jun 2001 P D Zamore T Tuschl P A Sharp and D P Bartel Rnai double stranded rna directs the atp dependent cleavage of mrna at 21 to 23 nucleotide intervals Cell 101 1 25 33 Mar 2000 97
7. as hits The genome matching reads are normalised and weighted by repetitiveness The normalisation method divides hit counts by the number of redundant reads that match the genome The normalised count for each distinct read is given in hits per 1 million matching reads Because it is impossible to decide where a sRNA with multiple matches to the genome originated we correct the normalised read abundance for repet itiveness by dividing it by the number of matches to the genome The result is a weighted hit count The method uses the normalised and weighted read abundance and relative position of sRNAs on the reference genome to predict the sRNA loci A locus must have a minimum of 3 weighted sRNA hits this threshold can be adjusted using the min_hits parameter and no gap absence of sRNA hits longer than 300nt this threshold can be adjusted using the max_gap parameter By default SiLoCo compares two sRNA samples by computing the loge sRNA expression ratio and the expression average These measures are used for ranking to help find differentially expressed loci Although SiLoCo compares two samples by default single sample mode can also be selected The datasets must contain sRNA sequence reads in FASTA format in redundant form i e with one entry for each read Sequences shorter than 18nt minsize parameter or longer than 30nt maxsize parameter will be removed Before finding sRNA loci we remove low complexity sequ
8. contain the same number of values which should be the expression levels of sRNAs genes in the series Each column should contain data coming from the same sample e g a time point or treatment The order of sRNAs genes should be identical for all columns forming the table The header row should indicate the name of each time point that will be analyzed The default format assumes the existence of a header row the first row will always be considered to be the header of the table Gene WildType Mutantl Mutant2 Mutan3 Mutant4 Mutant5 Genel 7 34 12 57 10 14 7 29 7 33 5 44 Gene2 5 32 5 53 5 12 10 45 10 39 10 47 Gene3 5 12 5 07 10 78 7 12 7 09 3 23 Gene4 6 54 14 58 9 19 6 89 7 03 5 12 Gene5 7 94 12 59 12 17 13 03 12 97 12 76 sRNA WildType Mutantl Mutant2 Mutan3 Mutant4 Mutant5 sRNA1 0 12 6 23 5 59 6 12 5 78 5 46 sRNA2 3 23 3 57 3 14 3 29 3 33 12 75 sRNA3 7 64 15 12 10 76 7 29 7 53 7 44 sRNA4 12 33 12 43 12 44 12 29 12 33 12 45 sRNA5 16 43 0 21 8 44 7 99 8 58 7 78 The input files can be generated by combining expression levels of the dif ferent products SRNA loci genes etc for different conditions points For example the weighted and normalised expression level of sRNA loci can be obtained using the SiLoCo tool in single sample mode Expression profiles for microarray datasets can be obtained using standard functions in packages such as affy in the Bioconductor
9. each too Default values are intended to filter heavily the sRNA reads input which reduces run time The general format for calling each tool is srna tools pl tool NAME_OF_THE_TOOL 0PTIONS_FOR_TOOL The help for each tool is available using the following command srna tools pl tool NAME_OF_THE_TOOL help 11 2 2 Example of a Bioinformatic Analysis Tutorial Given FASTA file files GSM118373_ Rajagopalan leaf fa a typical anal ysis would start with removing the adaptors If instead of a FASTA file a FASTQ file is provided the first step also included in the adaptor tool is converting the FASTQ file to FASTA format The adaptor removal tool is invoked with the following command srna tools pl tool adaptor adaptor_sequence_3 TCGT out output a srna_file files GSM118373_Rajagopalan_leaf fa This generates a file srnas_adapters_removed fa in output a Next to filter all sequences that do not map to the genome and are outside a certain size range the Filter tool is used The size range filter can be used for example to focus the analysis on putative miRNA candidates srna tools pl tool filter srna_file output a srnas_adapters_removed fa out output f genome data arabidopsis fa make_nr maxsize 26 minsize 16 The results of the filter tool are contained in a zip archive which can be unzipped using the command unzip filter_results zip in the directory output f The file MyJob_filtered fast
10. positive or negative is shown in the last column of the table If present annotations on both genes and sRNAs can facilitate a biological hypothesis Is the correlation coefficient reliable If the expression values in both series gene and sRNA series are comparable i e the expression ranges are comparable then the Pearson Correlation Coefficient will accurately compute the similarity between series 52 4 2 7 SiLoCo Some input sequences are excluded from analysis Sequences shorter than 18nt or longer than 30nt will be automatically re moved In addition we remove low complexity sequences that consist of one or two bases only such as AGAGAGAGAGAGAGA 4 2 8 SiLoMa How do I know the GMOD Genome Browser is working You will get an error on the command line if this is the case How can I represent the sRNAs in different samples Currently the multiple sample feature is not supported The input files can be either merged in one file and displayed or independent figures can be created for each sample 53 4 2 9 ta siRNA Prediction Tool How does noise influence the accuracy of the results Currently we do not apply any cleaning procedure by default However the data can be filtered using the Filter tool before using the ta si Prediction tool Some sequences are excluded from analysis Some input sequences are excluded from analysis e Only sequences with a read count of two or more are included e Only sRNAs of 21nt
11. ATATAGAACATACCGTTTTCCTTCTAGTTTTGTATATATAACCAAAATTAGTAG ACTTCAATTTTTC The following information is shown 1 sRNA ID accession 2 Target transcript ID accession and start end position of the target site 3 Any information annotation this sequence may have 4 Alignment of the miRNA bottom sequence to the target site top sequence represents a base pair represents a mismatch Uan o represents a G U basepair 5 Full sequence of the predicted target In addition a csv file containing a summary of all potential targets is pro duced 45 Chapter 4 Troubleshooting and FAQ 4 1 General What is a FASTQ file FASTQ format is a text based format for storing both a biological sequence usually nucleotide sequence and its corresponding quality scores Both the sequence letter and quality score are encoded with a single ASCII character for brevity Our tools do not use the quality scores in the FASTQ files only the sequence is used for downstream processing A FASTQ file normally uses four lines per sequence Line 1 begins with a character and is followed by a sequence identifier and an optional description like a FASTA title line Line 2 is the raw sequence letters Line 3 begins with a character and is optionally followed by the same sequence identifier and any description Line 4 encodes the quality values for the sequence in Line 2 and must contain the same number of
12. ME_reference sequence fasta is the reference transcript to which sRNAs were aligned JOBNAME_image png is the genome browser figure of the reference transcript and aligned sRNAs The ruler shows the distance along the reference transcript relative to its start The sRNAs are shown as arrows The direction of the arrows indicates the match to posi tive pointing right or negative pointing left strand and the colour indicates the sRNA size class see figure 3 8 pink 15 20nt red 20 21nt green 22 23nt blue 24 25nt 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 h A E aaa AROSA EEE EEE EEE HEE EEE EEE EEE EH EHH HEE HHHH HHHH 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 SO gt h SS Figure 3 8 SiLoMa output showing sRNAs with different lengths Note only sRNAs that have a full length perfect match to the reference transcript are displayed If labels are included some labels near the edges may be cut off To avoid this simply expand the region of interest on the reference transcript to allow the labels to be printed in full Also the arrows displaying the sRNAs may be larger than the region the sRNA maps to when 38 using the non redundant format with highly abundant sequences and a large region Examples of sRNA loci are presented in 3 9 and S1_flower_bud 2 openf Louer 3_fruit_1_to_3nn 54_fruit_5_to
13. The UEA sRNA Toolkit A User Guide for the Perl Implementation Martin Lott Daniel Mapleson Irina Mohorianu Vincent Moulton Simon Moxon Frank Schwach Contact srna tools cmp uea ac uk May 18 2012 Contents 2 5 1 1 Deliverablesl 5 O ant es Ge ea aa ee oe oe Se See 5 2 Quickstart 8 21 SUMMA le ode we a Oe ee Eo ee ye a od 8 Sane wi be 2 dw Poe hoe wee oe Be A 12 3_ Tools 15 3 1 Sequence File Pre Processing Tool 15 32 Filter To0ll a 17 3 3 mIRCat Tool s a si ere ee oR es ee ae a a 19 3 4 MiIRProf Tool occiso 24 3 5 RNA Hairpin Folding and Annotation Tool 28 3 6 FiRePat Tool a a sp au ia aa ew ee ee aa 29 3 SiLoCo LOOM r gg aea a e a a a ee a i i 32 3 8 SiLoMa Tool 36 3 9 ta siRNA Prediction Tool 41 3 10 Plant Target Prediction Tool 0 44 4 Troubleshooting and FAQ 46 Al General oos ss ieor aod Sow eR he Cae be a a 46 stds ceed Ge e e de ep We Bee ee 49 55 Preface RNA silencing is a complex highly conserved transcriptional and post transcriptonal mechanism that tunes gene expression It was originally studied as a defense mechanism against pathogens in plants 4 and later studied extensively due to its ability to regulate cancer related genes 29 RNA silencing is mediated by molecules known as small RNAs sRNAs which are reviewed in detail in 26 Recently high through
14. _7mn 58_fruit_breaker_plus_3 S9_fruit_breaker_plus_5 10_fruit_breaker_plus_7 y SALA Figure 3 9 miRNA locus 39 10000 10100 10200 10300 10400 10500 10600 10706 1_flowert bud aE De tel pi S2_open f Lower es E q S3_fruit_1_tg_3mn itt i Db bb aid 4_fruit_5_to_7nn e a S6_Fruit_nafure_gr a gt e y E 7 98_fruit_bry a S E aji e pd HR Figure 3 10 hcRNA locus 3 9 ta siRNA Prediction Tool This tool identifies phased 21nt sRNAs characteristic of ta siRNA loci It implements the algorithm described in 7 to calculate the probability of the phasing being significant based on the hypergeometric distribution see figure 3 11 Our implementation differs slightly as we take into account the length of the input sRNA sequences only using 21nt sRNAs in the phasing analysis We also require that sRNAs have a raw abundance of at least 2 in order to be included in the analysis A J f 21nt g 3 l 231bp 40 21 B k k n number of distinct small RN As identified in 231 bp region n k E Pr x k k number of distinct small RNAs mapped to phased positions S 2 P value p k Y Pr x pa Figure 3 11 prediction fo ta siRNA loci Parameters e Required genome The location of the genome file in FASTA format srna_file The location of the sRNA file in FASTA format out The path of the output directory e Optional abundance The mini
15. a contains the filtered sequence data in FASTA format Next to predict novel miRNA candidates miRCat can be used as follows srna tools pl tool mircat genome data arabidopsis fa srna_file files GSM118373_Rajagopalan_leaf fa out output m genomehits 5 hit_dist 100 maxgaps 4 max_overlap_length 50 max_percent_unpaired 60 max_unique_hits 4 maxsize 24 min_abundance 6 min_energy min_gc 20 min_hairpin_len 80 min_paired 25 minsize 19 no_complex_loops percent_orientation 80 pval 0 2 trrna window_length 100 miRCat identifies both old and new miRNAs and creates a zip archive con taining the annotations of each hairpin for the miRNA candidates struc tures pdf constructed using the hairpin annotation tool and the Vienna package 14 This example is for illustration purposes only The GSM118373_Rajagopalan leaf fa file contains Illumina sequences with adaptors removed and is already in FASTA format 12 10 0 Next to visualise other hairpins produced using miRCat the RNA folding tool can be used The input for this tool consists of the hairpin sequence and the miRNA miRNA which will be highlighted on the secondary structure srna tools pl tool hp_tool longSeq files hairpin fa shortSeqs files mirna fa out output h In order to determine the expression of each known miRNA in the sample miRProf can be used The results from miRProf are comparable across samples and hence miRProf is
16. acs instead The list is initially sorted by chromosome and position e Raw count S1 S2 Sum of read abundances in samples 1 and 2 that from the locus not corrected for repetitiveness e Weighted count S1 S2 Sum of raw read abundances divided by number of matches of each sequence to the genome e Normalised count S1 S2 Sum of weighted counts divided by the total number of genome matching reads in each sample given in hits per 1 million genome matching reads Normalised counts abundances are comparable between sam ples e Uniquely matching reads optional Number of sequence reads in the locus that only have a single match to the genome e log ratio A measure for the difference in sRNA abundance for a given locus between the two samples expressed as logy L When a locus is absent in one of the samples i e the expression level in one of the samples is 0 the ratio S1 S2 will be either 0 or inf It is not possible to calculate log2 0 or loga inf To avoid this problem a small pseudocount with default value of 0 1 is added to all normalised and weighted hit counts The bias introduced by the arbitrary pseudocount becomes negligible in loci with high expression levels A loga ratio of 1 means a two fold change in sRNA abundance A locus with a positive loge ratio shows an enrichment of sRNAs in samplel a locus with a negative ratio shows an enrichment in sample2 Unlike the linear ratio S1 S2 log ratios are symmetrical
17. ad and Rafael A Irizarry affy analysis of affymetrix genechip data at the probe level Bioinformatics 20 3 307 315 Feb 2004 Sam Griffiths Jones Alex Bateman Mhairi Marshall Ajay Khanna and Sean R Eddy Rfam an rna family database Nucleic Acids Res 31 1 439 441 Jan 2003 Gregory J Hannon Rna interference Nature 418 6894 244 251 Jul 2002 Ivo L Hofacker Rna secondary structure analysis using the vienna rna package Curr Protoc Bioinformatics Chapter 12 Unit12 2 Jun 2009 Ana Kozomara and Sam Griffiths Jones mirbase integrating microrna annotation and deep sequencing data Nucleic Acids Res 39 Database issue D152 D157 Jan 2011 Tamara Kulikova Philippe Aldebert Nicola Althorpe Wendy Baker Kirsty Bates Paul Browne Alexandra van den Broek Guy Cochrane Karyn Duggan Ruth Eberhardt Nadeem Faruque Maria Garcia Pastor Nicola Harte Carola Kanz Rasko Leinonen Quan Lin Vin cent Lombard Rodrigo Lopez Renato Mancuso Michelle McHale Francesco Nardone Ville Silventoinen Peter Stoehr Guenter Stoesser Mary Ann Tuli Katerina Tzouvara Robert Vaughan Dan Wu Weimin Zhu and Rolf Apweiler The embl nucleotide sequence database Nu cleic Acids Res 32 Database issue D27 D30 Jan 2004 Anthony A Millar and Peter M Waterhouse Plant and animal micror nas similarities and differences Funct Integr Genomics 5 3 129 135 Jul 2005 Attila Molnar Frank Schwach David J Studholme Eva C Thuene mann
18. alysis the most significant pairs a high similarity threshold should be given as input parameter The resulting pairs are sequentially clustered using two methods first by hierarchical clustering and then by k means The first method suggests a putative number of clusters and then an automated procedure selects the optimal number of clusters which is used in the k means clustering Parameters e Required gene_file The location of a gene expression file in csv format e g see example in files directory srna_file The location of a sRNA expression file in csv format e g see example in files directory out The path of the output directory e Optional 30 color_int Number of color intervals for html output 1 lt color int default color_int 10 de threshold Differential expression threshold 1 lt de_threshold lt 100 default de_threshold 5 Note that increasing this parameter will increase the number of selected series and thus the number of possible pairs slowing down the analysis sim threshold Similarity threshold 85 lt sim_threshold lt 100 default sim_threshold 95 Example srna tools pl tool firepat out output fp gene_file files firepat_test150_genes csv srna_file files firepat_test150_srna_loci csv color_int 3 de_threshold 30 sim_threshold 95 In order to emphasize the changes in expression the original data is trans formed to loga ratios relative to t
19. and David C Baulcombe mirnas control gene expression in the single cell alga chlamydomonas reinhardtii Nature 447 7148 1126 1129 Jun 2007 56 19 20 21 22 23 24 25 26 27 28 29 Ali Mortazavi Brian A Williams Kenneth McCue Lorian Schaeffer and Barbara Wold Mapping and quantifying mammalian transcrip tomes by rna seq Nat Methods 5 7 621 628 Jul 2008 Rebecca A Mosher Frank Schwach David Studholme and David C Baulcombe Polivb influences rna directed dna methylation indepen dently of its role in sirna biogenesis Proc Natl Acad Sci U S A 105 8 3145 3150 Feb 2008 Simon Moxon Frank Schwach Tamas Dalmay Dan Maclean David J Studholme and Vincent Moulton A toolkit for analysing large scale plant small rna datasets Bioinformatics 24 19 2252 2253 Oct 2008 Kay Prufer Udo Stenzel Michael Dannemann Richard E Green Michael Lachmann and Janet Kelso Patman rapid alignment of short sequences to large databases Bioinformatics 24 13 1530 1531 Jul 2008 Ramya Rajagopalan Herv Vaucheret Jerry Trejo and David P Bar tel A diverse and evolutionarily fluid set of micrornas in arabidopsis thaliana Genes Dev 20 24 3407 3425 Dec 2006 Brenda J Reinhart Earl G Weinstein Matthew W Rhoades Bon nie Bartel and David P Bartel Micrornas in plants Genes Dev 16 13 1616 1626 Jul 2002 Rebecca Schwab Javier F Palatnik Markus Riester Carla Schommer Markus
20. and EMBL 16 release 95 09 Jun 2008 The file can be replaced with any FASTA file containing t rRNAs sequences Note the tool might remove some sequences that are not t rRNA simply due to a random match to an annotated t rRNA in another species which are present in the file data t_and_r_RNAs fa Then if the user provides a corresponding genome the sequences can be partitioned into genome matching and not genome matching Usually the reads that do not map to the genome are considered sequencing errors or minor contamination and are generally discarded Another application for genome filtering is the analysis of reads produced from virus treatment ex periments For example these sRNA reads can be partitioned into three categories reads identified in both the host and viral genome reads unique only to the host genome and reads unique only to the viral genome This can be achieved by running the filter tool several times with the different genomes Parameters e Required srna_file The location of the sRNA file in FASTA format out The path to the output directory e Optional 17 genome A FASTA file containing a genome Sequences can be filtered according to whether they match or not the genome By default only sRNAs matching the genome are kept make_nr If specified the resulting FASTA file is made non redundant The file will be smaller as there is only one entry per unique se quence Do not use this option i
21. are included e Only 21nt phase groups are identified e Low complexity sequences those composed of fewer than three dif ferent nucleotides are filtered out to limit the size of genomic match files 4 2 10 Plant Target Prediction Tool Can not find transcriptome Some transcriptomes can be found in INSTALL_PATH data transcriptomes and referred to only by the file name on the command line e g transcriptome arabidopsis fa will cause the program to look for a transcriptome at INSTALL_PATH data transcriptomes arabidopsis fa 54 Bibliography 1 N w Edwards Allen Zhixin Xie Adam M Gustafson and James C Carring ton microrna directed phasing during trans acting sirna biogenesis in plants Cell 121 2 207 221 Apr 2005 Tyler W H Backman Christopher M Sullivan Jason S Cumbie Zachary A Miller Elisabeth J Chapman Noah Fahlgren Scott A Gi van James C Carrington and Kristin D Kasschau Update of asrp the arabidopsis small rna project database Nucleic Acids Res 36 Database issue D982 D985 Jan 2008 Tanya Barrett Tugba O Suzek Dennis B Troup Stephen E Wil hite Wing Chi Ngau Pierre Ledoux Dmitry Rudnev Alex E Lash Wataru Fujibuchi and Ron Edgar Ncbi geo mining millions of ex pression profiles database and tools Nucleic Acids Res 33 Database issue D562 D566 Jan 2005 D C Baulcombe In vitro replication of plant viral rna Curr Biol 1 1 53 54 Feb 1991 James C Carrington and V
22. around zero 34 Note an increased sRNA abundance in one sample does not necessarily mean that sRNA expression from that locus is upregulated Consider the case of a mutant that looses expression of sRNAs from all but a few loci These loci will show an increased sRNA abundance compared to the wild type because other sRNAs are missing but sRNAs could still be produced at the same rate from these loci in vivo In order to rank the loci we use the following measures e average normalised count The loga ratio alone is not sufficient for finding differentially expressed loci because this measure is unreliable when the sRNA abundance is low in both samples Good candidate loci should have a high ratio of sRNA abundance and a high average count e loga ratio rank Each locus is given a rank according to its absolute loga ratio Low rank numbers indicate a high degree of enrichment depletion Equal logz ratios share a rank e average based rank Is similar to the logg ratio rank but based on the average normalised counts e weighted rank sum This measure can be used to identify candidate loci that show a high degree of enrichment depletion in one of the samples at a high overall expression level The rank sum is calculated as follows RS 0 5x RR 0 5 x AR where RS is the rank sum RR is loge ratio rank and AR is the average based rank Example output SiLoCo can be used to compare two Gene Expression Omnibus GEO datas
23. e and the thickness of the arrow is proportional to the logio of the sequence abundance If plot_nr is not present redundant output is requested multiple arrows are drawn to represent the abundance of each sequence in the sRNA file Figure 3 7 shows the differences created by the plot_nr parameter Labels plot_labels can also be included in the output which contain the sRNA sequence and its abundance 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 AAA o AA AAA Figure 3 7 SiLoMa output in redundant A and non redundant B format After processing SiLoMa produces an archive that contains the following files e JOBNAME stats txt a table in CSV format containing an overview of the number of matches total and unique sequences to the reference transcript For example for a sRNA with abundance 2 which matches 4 times to 37 the reference transcript the number of sequences displayed in redun dant mode will be 1 x 2 x 4 8 The number of sequences displayed in non redundant mode will be 1 x 4 4 see example presented in figure 3 7 In addition there is also a break down by strand and sRNA size class see figure 3 8 JOBNAME_matches fasta is a FASTA formatted list of the matching sequences in alphabetical order The IDs are in the following format CONSECUTIVE NUMBER_COUNTx_pos MATCH POSITIONS where MATCH POSITIONS are in the format START END and represents the strand JOBNA
24. e t rRNA file contains sequences from Rfam 10 the Genomic tRNA Database 6 and EMBL 16 release 95 09 Jun 2008 and might be out of date This file can be replaced with a more recent file downloaded from http www sanger ac uk resources databases rfan html The sequence of interest might have a random match to an annotated t rRNA in another species If you are not sure what sequences should be kept for further analysis and which sequences should be removed we suggest you to leave the t rRNA filtering for later steps When should I use the make_nr option This option prepares a FASTA file for e g complexity analysis number of unique sequences to number of redundant sequences It also represents a compact version of the redundant file making it a suit able solution for data storage However some tools require the input FASTA file in redundant format Please check the specifications for each tool prior to running it 49 4 2 3 miRCat Does the tool predict all known miRNAs present in the sample Possibly not since miRNAs present at low abundance may be filtered out Which are the best miRCat candidates The best indication of a good miRNA is a high abundance and presence of a miRNA The randfold p Value also provides some indication of the quality of the miRNA 4 2 4 miRProf miRBase files not prepared for overhanging matches miRBase files are currently not prepared for overhanging matches if you email us we will do o
25. ences and matches to known t rRNAs Parameters e Required genome The location of the genome file in FASTA format sample namel The name of the first sRNA sample e g S1 srna_filel The location of the first sRNA FASTA file out The path of the output directory e Optional 32 sample name2 The name of the second sRNA sample e g S2 required if num_samples 2 srna_file2 The location of the second sRNA FASTA file required if num_samples 2 asrp_links If defined ASRP links to Arabidopsis small RNA database ASRP 2 will be added to the results file max_gap The maximum gap length in a locus 1 lt maz_gap default max_gap 300 maxsize The maximum length of a sRNA 18 lt mazxsize lt 35 default maxsize 25 min_hits The minimum number of sRNAs in a locus 1 lt min_hits default min_hits 3 minsize The minimum length of a sRNA 18 lt minsize lt 35 default minsize 18 num samples The number of samples num_samples 1 or num_samples 2 default num_samples 2 pseudocount The pseudocount that is added to locus expression level to avoid division by zero errors 0 lt pseudocount default pseudocount 0 1 tair_links If defined links to TATR 27 will be added to the results file Links work in MS Excel and OpenOffice calc but there may be versions of these programs with which they do not work Please note that the hyperlinks will increase the size of your resul
26. ent_unpaired The maximum percentage of unpaired bases in hairpin 1 lt max_percent_unpaired lt 100 default max_percent_unpaired 50 max_unique_hits The Maximum number of non overlapping hits ina locus 1 lt max_unique_hits default max_unique_hits 3 maxsize The maximum length of a miRNA 18 lt mazsize lt 24 default maxsize 22 min_abundance The minimum sRNA abundance 1 lt min_abundance default min_abundanc 5 min energy The minimum free energy of the hairpin Must be lt 0 Default 25 min_gc The Minimum percentage of G C in miRNA must be gt 1 and lt 100 Default 10 min hairpin len The Minmum length of hairpin nt must be gt 50 Default 75 min_paired The Minimum number of paired bases in miRNA re gion Must be gt 10 and lt 25 Default 17 minsize The Minimum sRNA size Must be gt 18 and lt 24 De fault 20 no_complex_loops If defined the hairpins with complex loops are removed percent orientation The percentage of sRNAs in locus that must be in the same orientation 1 lt percent_orientation lt 100 de fault percent_orientation 90 pval The p value 0 0 lt pval lt 1 0 default 0 1 trrna If defined sRNAs matching sequences in the FASTA t r RNA file data t_and_r_RNAs fa will be removed window_length The window length 40 lt window_length lt 400 default window length 150 21
27. erl Packages The toolkit is available for machines running a Linux distribution For Debian based distributions the following packages and their dependencies are required e bioperl e libtemplate perl e libconfig auto perl e libexception class trycatch perl e libmail sendmail perl e libyaml tiny perl On Debian distributions for example on Ubuntu Linux BioPerl and its dependencies can be installed with the command sudo apt get install bioperl For non Debian distributions it is possible to build equivalent packages con tact with your local systems administrator 1 2 2 Required Binaries In addition to the Perl package dependencies the toolkit requires a number of executable programs such as PatMaN and Vienna 14 These are provided in the srna tools usr local bin zip archive which should be extracted and copied to for example usr local bin or alternatively to some location pointed to by your path variables If you do not have root permissions please contact your local systems administrator 1 2 3 Configuration Before using the toolkit you must edit the configuration file INSTALL_PATH config application conf and insert the full path of the directory where the toolkit resides on your system see figure L 1 This will enable the different parts of the toolkit to find each other 56 FEFSESSE FSS S SSS SEE ESSE S SESS ESS SSS FSS ESSE SESE 57 Paths to directories on server and cluster 58 where we put jobs in
28. ets such as the Arabidopsis flower and leaf sets from the Bartel lab a e Flower sample INSTALL_PATH files GSM118372 Rajagopalan_col0_flower fa e Leaf sample SINSTALL PATH files GSM118373_Rajagopalan col0 leaf fa This file has already been sorted by rank sum and links to the ASRPdb genome browser are included The top ranking loci in this analysis show examples of loci that are highly differentially expressed in leaf and flower tissues 35 3 8 SiLoMa Tool This tool produces a map of sRNAs that match to a reference transcriptome using GMOD Generic Model Organism Database project http gmod org wiki GMOD genome browser Each sRNA like read is shown using a colored arrow to indicate the precise location orientation and abundance with respect to the transcriptome see figure 3 6 118k 115k 120k s4 3_fruit_i_to_3nn 3 e gt gt les gt 4_fruit_5_to_7nn SOAGTCTCTEATCGTAGATEGT CAACARAGAACCATACACTGAA AAGAAACTGAAGACGCTTTCCT TrocacraTaGTCcATAACARAC TGGACTTATAAGCACTIGACTO AACARAGAACCATACACTGART TGGTAAGARACTGARGACGCTT CTOGACAGCTTCSGATGGACAA gt GTGGTAAGARACTGAAGACEC TETGOTRAGARACTORAGACEO Figure 3 6 SiLoMa output showing a compact sRNA locus The reference sequences forming the transcriptome are comprised of either a region in a user supplied genome or a user supplied sequence To match against a region within a genome enter a chromosome BAC scaffold ID e g one of 1 2 3 4 or 5 for Arabidops
29. f the pre dicted TAS loci along with their abundances and genomic coordinates For the example shown below Chr Start postition End position sequences phased sequences p val 2 16544875 16545126 13 8 4 833452e 09 the ta siRNAs are Read Chromosome Start position Strand CCAATGTCTTTTCTAGTTCGT 19 2 16544875 1 CGCTATGTTGGACTTAGAATA 6 2 16544917 1 ATTTTCTAAGATCCACCGATA 12 2 16544938 1 GAACTAGAAAAGACATTGGAC 4 2 16544893 1 TTCTAAGTTCAACATATCGAC 12 2 16544914 1 TTCTAAGTCCAACATAGCGTA 301 2 16544935 1 TCGGTGGATCTTAGAAAATTA 161 2 16544956 1 TACAAGCGAATAGACCATTTA 12 2 16544977 1 42 The first line diplays the locus coordinates as shown in the previous file Sub sequent lines show the ta siRNA sequences with the abundance in brackets e g TICTAAGTCCAACATAGCGTA 301 The sequence coordinates chromo some start position orientation are also shown for each of the predicted ta siRNAs The tool has been tested using the sRNA set in files GSM118373_Rajagopalan_leaf fa described in 23 The results obtained using default parameters are shown below Chr Start End seqs phased seqs p val LocusInfo 1 18553086 18553337 16 9 1 183951e 09 TAS1b 1 23305788 23306039 6 4 4 549688e 05 PPR repeat gene 2 11729024 11729275 27 10 4 833452e 09 TASla 2 16544875 16545126 13 8 4 833452e 09 TASIc 2 16546892 16547143 29 11 1 886064e 09 TAS2 3 1970346 1970597 5 4 1 563104e 05 AT3G06435 1
30. f you wish to use the filtered list with other tools in this toolkit maxsize The maximum length of a read 16 lt mazsize lt 35 default marsize 25 minsize The minimum length of a read 16 lt maxrsize lt 35 default minsize 18 trrna If defined all reads matching a sequence in INSTALL_PATH data t_and_r_RNAs fa are removed trrna_sense By default both sense and anti sense t rRNA matches are accepted If defined only sense matches are removed discard genome matching Rather than keeping genome matches only sequences that don t match the genome are retained Example srna tools pl tool filter srna_file files GSM118373_Rajagopalan_leaf fa out output f genome data arabidopsis fa make_nr maxsize 26 minsize 16 trrna trrna_sense 18 3 3 miRCat Tool miRNAs are a well studied class of sRNAs 24 that are generated from a single stranded RNA ssRNA that forms a stable partially double stranded stem loop structure hairpin 9 miRCat predicts miRNAs from high throughput sRNA sequencing data without requiring a putative precursor sequence as these will be identified by the program The tool receives as input two FASTA files the sRNA sequence file with adaptors removed and a corresponding genome for the organism that is being studied Before processing miRCat maps the sRNA sequences to the genome using PatMaN 22 PatMaN is provided in the dependencies archive for the tool
31. file containing predicted miRNA precursor structures with miRNA and miRNA if present highlighted 22 e A FASTA format file of all predicted mature miRNA sequences Suggested parameters for the animal and plant version of miRCat are listed below Parameter Plant Animal window_length 100 AO min_paired 17 17 min_abundance 5 20 5 20 max_gaps 3 3 max_genome hits 16 16 min_length 20 21 max_length 22 23 min_hairpin_len 75 50 hit_dist 200 50 pval 0 1 0 1 no_complex_loops false true max_unpaired 60 40 orientation 80 80 The following parameters can be left as default minenergy max_overlap_length min_gc 23 3 4 miRProf Tool This tool determines the expression levels of sRNAs that match known miR NAs The expression level of a sRNA represents the number of occurences of the sequence in the sample miRProf allows the user to group miRNAs according to different criteria e g organisms and or family miRProf filters the sequences before the expression level is computed Se quences shorter than 18nt min_size or longer than 30nt max_size will be removed In addition low complexity sequences that consist of one or two bases such as AGAGAGAGAGAGAGA are removed The user also has the option to filter against t rRNA and a user specified genome Filtering will have an impact on the number of reads used for normalisation After building the expression le
32. gend Parameters e Required longSeq The long hairpin sequence in FASTA format use quotes to give a parameter on more than one line shortSeqs The short sequence s in FASTA format that will be highlighted on the hairpin These should be subsequences of the longSeq out The path of the output directory Example srna tools pl tool hp_tool longSeq files hairpin fa shortSeqs files mirna fa out output h miR164 ER A a DEFOE POCOO a Pox Seed re ORENSE ESPOSO 2 Lobe coco SPOR ECR a miR164 Figure 3 4 RNA fold output showing miR164 precursor 28 3 6 FiRePat Tool This tool identifies sRNAs or sRNA loci that may influence gene expres sion To do this FiRePat Find Regulatory Patterns computes the profile similarity between series of sRNA and gene expression data Pairs of enti ties SRNAs genes that are highly co or anti regulated are identified and optimally clustered In both cases the Pearson Correlation Coefficient is used as a similarity measure Gene sRNA profiles with a high degree of co or anti regulation might indicate a functional interaction and are therefore interesting subjects for further studies 3 6 1 Input files In order to use FiRePat you need expression profiles of sRNAs and genes in at least two samples e g a time series different treatments or mutants The input of this tool consists of two CSV files containing the series Each row in the input file should
33. gnificant based on a hypergeometric distribution 7 10 Plant target prediction tool by S Moxon Using a FASTA file containing sRNAs in non redundant form and a FASTA file containing pairs SRNA transcript are predicted based on the rules suggested in I and 25 Both plants and animals use sRNAs to regulate gene expression How ever some sRNA types may not be present in both plants and animals e g trans acting short interfering RNAs ta siRNAs are plant specific siRNAs and piRNAs are animal specific siRNAs Also some sRNA types such as microRNA miRNAs while having a similar biogenesis adjust the gene expression in slightly different ways in plants and animals 5 17 For this reason some tools in the toolkit are specific for plant data sets as shown in the following table Tool Animal Data sets Plant Data sets Pre processing Filter miRCat miRProf RNA folding FiRePat SiLoCo SiLoMa ta siRNA prediction Plant target prediction P A A A AAA In addition to specifying the input files and output directories you need the following Tool Sample Data Species Data Pre processing 1x FASTA sRNA file N A Filter 1x FASTA sRNA file 1x FASTA Genome miRCat 1x FASTA sRNA file 1x FASTA Genome miRProf 1x FASTA sRNA file 1x miRBase DB name RNA folding 1x FASTA sRNA file 1x FASTA Sequence FiRePat 2x CSV files
34. he first point A positive value across the newly created series suggests an increase in expression level relative to the first point and a negative value suggests a decrease in expression level The output consists of two csv files containing the positively and negatively correlated pairs respectively and two html files with a colored version of the tables see figure 3 5 The last two columns in the output files represent the correlation coefficient between the series that form the pair and the identification number of the cluster to which the pair belongs to C05HBa0028M20 1 81638 JEJ 82651 C07HBa0184E04 2 85812 88847 Les 3646 1 S1_at Les 3740 1 S1_at Les 3741 1 S1_at C08SLm0015J19 1 93121 9440 C12HBa0150C12 1 59342 Les 3756 1 S1_a_at nas C02HBa0040B13 1 111591 Les 4038 1 S1_at ES ma Les 4317 1 S1_at C03HBa0143N09 1 98718 100854 C02HBa0011A02 3 128523 128956 C06HBa0034C13 1 99998 100942 Les 4488 1 S1_at Les 764 1 S1_at C04SLm0130G07 1 23666 Les 764 2 A1_at Les 840 1 A1_at Figure 3 5 Firepat output on series containing 10 points 31 3 7 SiLoCo Tool This tool predicts sRNA loci using the method described in and 20 It also enables the user to compare the expression profile of sRNA loci between different samples In order to determine the relative position of sRNAs the reads are mapped to the reference genome using PatMaN 22 Only full length perfect matches are accepted
35. ictor Ambros Role of micrornas in plant and animal development Science 301 5631 336 338 Jul 2003 Patricia P Chan and Todd M Lowe Gtrnadb a database of transfer rna genes detected in genomic sequence Nucleic Acids Res 37 Database issue D93 D97 Jan 2009 Ho Ming Chen Yi Hang Li and Shu Hsing Wu Bioinformatic pre diction and experimental validation of a microrna directed tandem trans acting sirna cascade in arabidopsis Proc Natl Acad Sci U S A 104 9 3318 3323 Feb 2007 Noah Fahlgren Christopher M Sullivan Kristin D Kasschau Elisa beth J Chapman Jason S Cumbie Taiowa A Montgomery Sunny D Gilbert Mark Dasenko Tyler W H Backman Scott A Givan and James C Carrington Computational and analytical framework for small rna profiling by high throughput sequencing RNA 15 5 992 1002 May 2009 55 9 Marc R Friedlaender Wei Chen Catherine Adamidi Jonas Maaskola 10 12 13 14 15 16 17 18 Ralf Einspanier Signe Knespel and Nikolaus Rajewsky Discovering micrornas from deep sequencing data using mirdeep Nat Biotechnol 26 4 407 415 Apr 2008 Paul P Gardner Jennifer Daub John G Tate Eric P Nawrocki Diana L Kolbe Stinus Lindgreen Adam C Wilkinson Robert D Finn Sam Griffiths Jones Sean R Eddy and Alex Bateman Rfam updates to the rna families database Nucleic Acids Res 37 Database issue D136 D140 Jan 2009 Laurent Gautier Leslie Cope Benjamin M Bolst
36. is and a start and an end position The maximum length of the selected region is 50kbp Parameters e Required Either pasted_seq the reference transcript or genome the loca tion of the genome file in FASTA format srna file The location of the sRNA file in FASTA format out The path of the output directory e Optional maxsize The maximum length of a sRNA 16 lt mazsize lt 35 default maxsize 30 minsize The minimum length of a sRNA 16 lt minsize lt 35 default minsize 18 36 plot_labels Plot labels SRNA sequences and counts plot_nr Plot sRNA hits in non redundant form region chrom The chromosome of the reference transcript in the genome region start The start position of the reference transcript in the genome region_end The end position of the reference transcript in the genome Note if a genome is supplied only one transcript is created using the optional parameters region_chrom region_start and region_end Example srna tools pl tool siloma genome data arabidopsis fa out output sm srna_file files GSM118373_Rajagopalan_leaf fa pasted_seq TAAGCTATATAGGGGGGT region_chrom 2 region_start 39148 region_end 39445 maxsize 26 minsize 20 plot_labels plot_nr Some parameters like plot_nr control the graphical output of the tool If sequences are plotted in non redundant form only one arrow is drawn for each unique sRNA sequenc
37. kit Once the sequences are mapped to the input genome miRCat will look for genomic regions covered with sRNAs sRNA loci containing reads with abundance at least five this threshold can be adjusted using the min_abundance parameter These loci must match certain criteria see figure e Loci must contain no more than four non overlapping sRNAs e Each sRNA in a locus must be no more than 200nt away from it s closest neighbor this threshold can be adjusted using the hit_dist parameter e At least 90 of sRNAs in a locus must have the same orientation this threshold can be adjusted using the percent_orientation param eter A a 10 20 30 40 50 60 70 80 90 100 110 Figure 3 2 SiLoMa output showing miR164 Once a list of loci has been produced they are further analyzed in order to find likely miRNA candidates e The most abundant sRNA read within a locus is chosen as the likely miRNA e Flanking sequences surrounding this sRNA are extracted from the genome using varying window lengths 19 e Each sequence window is then folded using RNAfold producing a sec ondary structure for the putative miRNA see figure 3 3 e miRCat then trims the secondary structure and computes discrimina tive features useful for classifying miRNAs The features are The number of consecutive mismatches between miRNA and miRNA must be no more than 3 The number of paired nucleotides between the miRNA and the miRNA
38. lue is automatically assigned e g min_energy 12 5 e boolean parameters have default value 1 TRUE if they are required parameters and 0 FALSE if they are optional parameters e g trrna true If the input type does not match the parameter type an error is produced and the execution of the script is halted e g minsize numeric parameter minsize AGTC ERROR However in the case of a numeric mismatch i e real value instead of integer value displays no warning all values are rounded up to nearest integer number e g 18 1 19 The stand alone sRNA toolkit is a Unix based only product Therefore directory paths in this guide use a forward slash character as the separator between a directory name and the name of a subdirectory or file in that directory For example the absolute path arabidopsis srna_reads indicates the srna_reads subdirectory of a directory named arabidopsis mounted off the root directory on the file system Disclaimer The UEA bioinformatics group is not able to offer support for this ver sion of the toolkit However a new version will be released that will offer enhancements such as additional tools platform independence improved performance reduced hardware requirements and improved usability The UEA sRNA toolkit is free open source software distributed under the GNU General Public License Therefore the program is distributed WITH OUT ANY WARRANTY See the GNU General Public Lice
39. mbers and text are stored in plain textual form that can be read in a text editor Lines in the text file represent rows of a table and commas in a line separate what are fields in the tables row CSV files can be opened in MS Excel or other spreadsheet programs sRNA sequence Samplel Sample 2 Sample3 Sample4 AAAGTCGTA 10 Ts 100 25 GCTTCGAAA 100 10 20 55 GTCAGCTCC 34 7 25 53 CCGTAGCCA 37 2 64 67 ACGTCAGAG 27 5 1000 36 I pasted in the example code and it did not work First check if everything has been installed correctly In particular make sure the tool and the dependencies are available in the PATH Potentially the problem lies with the new line character You must remove the new line character before pasting the command e g paste into a text editor of your choice first remove then new line and then paste into the command prompt Cannot find error template file If you receive an error e g cannot find error template file that is because you have not set the correct paths in the application conf file patman bin file not accessible or not executable Make sure the PatMaN executable is in usr local bin or if not on the path for your local machine Use which patman to see if PatMaN can be found on the path Use chmod to give executable permissions Can not find organisms txt when updating miRBase This is normally caused by a problem writing to the data directory Ensure you have write access
40. mum sRNA abundance 1 lt abundance default abundance 2 pval The p value cutoff can be adjusted to increase decrease the number of loci returned Must be either 0 001 0 0001 0 00001 0 000001 or 0 0000001 de fault is 0 0001 minsize The minimum length of a sRNA 18 lt minsize lt 35 default minsize 18 maxsize The maximum length of a sRNA 18 lt marsize lt 35 default maxsize 25 Al trrna If defined sRNAs matching sequences in the FASTA t r Example RNA file data t_and_r_RNAs fa will be removed srna tools pl tool phasing genome data arabidopsis fa out output p srna_file files GSM118373_Rajagopalan_leaf fa abundance 3 pval 0 001 minsize 20 maxsize 26 trrna The results consist of two files The locuslist csv file contains a list of predicted TAS loci in csv format which contains the following information Chr Start postition End position sequences phased sequences p val 1 18553086 18553337 16 9 1 18e 09 e Chr Chromosome e Start position Start position of the ta siRNA locus e End position End position of the ta siRNA locus e sequences Number of unique sRNAs mapping to this locus e phased sequences Number of unique sRNAs in phase e p val p Value showing the probability of the phasing occurring by chance The srnas txt file contains a list of phased sRNAs from each o
41. n specially prepped to do so the _plusX variant of the database If not defined sRNAs with overhanging bases are always rejected For example this would be counted as 2 mismatches sRNA TTAAACCTAGGCAAATAACGATG PEEL E EEE EE dd dolo miRNA TTAAACCTAGGCAAATAACGGT trrna If defined sRNAs will be removed from the analysis if they match sequences in the FASTA t r RNA file data t_and_r_RNAs fa There are known miRNAs in some species that have a perfect match to the other genomes but are not yet annotated as miRNAs on the newer genomes To view matches to known miRNAs from a specific organism only you should not use the group_organisms option Example srna tools pl tool mirprof mirbase_db plant_mature out output mp srna_file files GSM118373_Rajagopalan_leaf fa collapse_match_groups genome data arabidopsis fa group_family 26 group_mismatches group_organisms group_variant keep_best maxsize 26 minsize 16 mismatches 2 overhangs trrna 27 3 5 RNA Hairpin Folding and Annotation Tool This tool produces the secondary structure of a long up to 1kb RNA sequence and annotates it by highlighting up to 20 short sequences on the resulting structure The tool produces three files e PDF file showing the position of miRNA candidate sequences on a precursor hairpin see figure 3 4 e JPEG file showing the position of miRNA candidate sequences on a precursor hairpin e A text file containing the le
42. nimum length of the read 16 lt minsize lt 35 default minsize 18 maxsize The maximum length of a read 16 lt mazxsize lt 35 default maxsize 25 Example srna tools pl tool adaptor adaptor_sequence_3 TCGT srna_file files GSM118373_Rajagopalan_leaf fa out output a adaptor_sequence_5 TGGA allow_rev_comp minsize 20 maxsize 25 16 3 2 Filter Tool This tool filters sRNA sequence files in FASTA format according to user defined criteria It generates a FASTA file with sequences that passed the filter s In addition a comma separated values csv table is produced which summarises the total number of sequences after each filtering step and the distribution of their lengths The sequences can be filtered based on their length using the optional pa rameters minsize and maxsize This will clean the input and prepare the reads for subsequent steps like miRNA prediction Next the low complexity sequences are filtered out This tool defines a sequence as having low complexity if it contains at most two distinct nu cleotides In addition the tool can filter transfer and ribosomal RNAs t rRNAs us ing the sequences present in the INSTALL_PATH data t_and_r_RNAs fa file This filtering is commonly conducted on sRNA datasets since reads map ping to tRNA and rRNA might be degradation products The file contains t rRNAs obtained from RFAM version 10 Jan 2010 12 0 the Genomic tRNA Database 6
43. normally run once on each sample and then the results are combined srna tools pl tool mirprof mirbase_db plant_mature out output mp srna_file output f MyJob_filtered fasta keep_best maxsize 26 minsize 16 mismatches 2 Finally the targets for the new and old miRNAs can be checked using the target prediction tool srna tools pl tool target out output t pasted_srnas gt a GCTTCTATCTTTTTCTTTCGTGCT transcriptome arabidopsis fa Besides identifying miRNAs we can identify all possible sRNA loci using SiLoCo and visualise the read distributions using SiLoMa srna tools pl tool siloco genome data arabidopsis fa out output si sample_name1 S1 sample_name2 S2 srna_filel files GSM118373_Rajagopalan_leaf fa srna_file2 files GSM154370_Carrington_col0_leaf fa srna tools pl tool siloma genome data arabidopsis fa out output sm srna_file files GSM118373_Rajagopalan_leaf fa pasted_seq TAAGCTATATAGGGGGGT region_chrom 2 region_start 39148 region_end 39445 After visually inspecting few genome browser figures we may wish to deter mine the ta siRNA loci present in a given plant dataset We can identify these loci using the ta siRNA prediction tool srna tools pl tool phasing genome data arabidopsis fa out output p srna_file files GSM118373_Rajagopalan_leaf fa Using MirProf we have obtained the expression profiles of the known miR NAs in the set After using SiLoCo we also have the ex
44. nse for more details a copy of which is available in the root directory of the software package and on the web at nttp www gnu org licenses gpl html Acknowledgements The sRNA toolkit was developed with support from the Biotechnology and Biological Sciences Research Council BBSRC http www bbsrc ac uk grants BB E004091 1 and BB 1I00016X 1 and the SIROCCO consortium http www sirocco project eu Chapter 1 Installation The toolkit can be downloaded and deployed onto machines running a Linux distribution The deliverables simply need to be unpacked into a directory of the user s choosing denoted in this document as INSTALL_PATH The user may find it helpful to ensure that the srna tools pl perl script is on the path In addition there are a number of dependencies that must be properly installed onto the system for the toolkit to function properly The remainder of this chapter describes the deliverables provided by the UEA as well as the toolkit s dependencies that must be installed onto the system 1 1 Deliverables The toolkit is split into three archives that can be downloaded from srna workbench uea ac uk perl_main_page html and are also collectively available as a CD iso image The archive files are names as follows srna tools cli zip The sRNA toolkit software srna tools usr local bin zip Software dependencies srna tools example zip Example files srna tools iso CD iso image 1 2 Dependencies 1 2 1 P
45. ocess unknown LOCATED IN chloroplast EXPRESSED IN 24 plant structures EXPRESSED DURING 15 growth stages BEST Arabidopsis thaliana protein match is SHW1 SHORT HYPOCOTYL IN WHITE LIGHT1 TAIR AT1G69935 1 Has 20 Blast 44 hits to 20 proteins in 5 species Archae 0 Bacteria 0 Metazoa 0 Fungi 0 Plants 20 Viruses 0 Other Eukaryotes 0 source NCBI BLink chr4 16201831 16203641 REVERSE 5 AGAAGAUGAUGAUGAUCACG AGGAAGAAGAUAGAAGCUUG 3 TIT Tolllottlittttttt 3 UCGUGCUUUCUUUUUCUAUCUUCG be gt AT4G33780 1 GAGCGTGTTGATGCATAACGAACGATGCCATTTTCCGCATCAATCTCATCGCCTTCTTCTTCTGTCGCG CTTCTTCGATCGCCTCTCTCTTTCTTCATCTTCACTCCCAAAACCCTAATCTTCACCAGAACCAGGATC TCTGGTTTCCCTTATCTTGCTTCCCGGCGATCCCGCGATTTCATCAACGGGAGGGATGATTTCGCTGAC GATACGAGGAGCTGGAACCGGAAGATCAAACCGGAGTATGGGTTCGATGAGGATTACGATGGAGAAGAA GATGATGATGATCACGAGGAAGAAGATAGAAGCTTGGATCTGTTACTTAGATTTGTAGAAAATGTTTTC AGAAAGATTTCTAAGAGAGCAAGGAAAGCTGTCCGATCAATTTTGCCTGTTTCGATCTCTACGAAGCTC GTGGGGTTTTCAGTGAATGGAGTACTTATTCTTGCTTTTTTGTGGATTTTGAAGGCTTTCCTCGAGGTA GCTTGCACACTTGGAACTATTGTATTTACGAGCATTCTACTTATACGTGGACTTTGGGCCGGAGTAGCA TACATGCAAGAGAGCCGCAACAATAGGATCAATGAACTCGCTGATGATCCTCGTGCATGGAACGGGATG CAACCAGTTTCCTGATGAATTCGCTTTACACTTGTAGAAATCAGAATTCTGACTTTTGGGAGAGCCATA ATTGTTTAGGTTCTTCCAAGGCAATAAAACCACAGCTGAGTTCAGAATCAGAAAGCAGTTACAGTGGAT GTTCATTGGCAATGTCTGATGATTTAGTAAGTAAAAAAAGTGTAATATTGTAGCATTCACCAAGTCAGC TATGCTGGTGTGTAGCTCAACTGGGAACTAAGTCGTCGCCAATGGTGACCATGTTTTCTTAGTTTCTAA ATAAATAAACCAAAC
46. on Predicts new miRNAs from high throuput sRNA sequencing data pre sented as a redundant FASTA file miRProf tool known miRNA expression profiler by F Schwach Determines the expression profile of sRNAs from a non redundant FASTA file that match known miRNAs from miRBase 15 RNA hairpin folding and annotation tool by F Schwach Produces the secondary structure of a long RNA sequence and an notates it by highlighting up to 20 short sequences on the resulting structure High Level Tools used for in depth data analysis 6 FiRePat tool Finding Regulatory Patterns by I Mohorianu Identifies positively and negatively correlated expression profiles of sRNAs sRNA producing loci and genes Receives as input two CSV files containg expression values in different samples 7 SiLoCo tool siRNA locus comparison by F Schwach Finds genomic sRNA producing loci by abundance and relative posi tion of sRNAs mapped to the reference genome 18 The sRNA files are required in FASTA format redundant form 8 SiLoMa tool siRNA locus mapper by F Schwach Maps sRNAs input given in FASTA format redundant form to a reference sequence and produces a genome browser image The loca tion and strand of each sRNA is represented with an arrow and the abundance of the sRNA is proportional to the thickness of the arrow 9 ta siRNA prediction tool by S Moxon Identifies ta siRNA loci by computing the probability of phasing being si
47. pression levels of the loci in at most two samples If similar gene data is available expression levels of genes measured in similar conditions we may use FiRePat to iden tify co anti regulated pairs using both the miRNA expression levels and the loci expression levels 13 srna tools pl tool firepat out output fp gene_file files firepat_test150_genes csv srna_file files firepat_test150_srna_loci csv The co and anti correlated pairs formed with miRNAs will help us decide which of the targets predicted by the target prediction tool are more likely to be real and these targets can be later validated in biological experiments The co and anti regulated pairs formed with loci will provide a general overview of interactions between sRNA loci and gene at genome level 14 Chapter 3 Tools 3 1 Sequence File Pre Processing Tool Sequencing devices produce reads with adaptor sequences at either end of the read This tool removes those adaptor sequences making the input file ready for use by other tools in the toolkit The tool is able to process a FASTQ or a FASTA file If a FASTQ file as produced by a sequencing device is provided as input then this tool first converts it to FASTA before the adaptors are removed It can also handle zipped and gzipped archives containing files of the above mentioned formats Next 5 optional and 3 required adaptors are removed as specified be low The 5 adaptor is optional beca
48. put sequencing has revolutionised the field of sRNA biology by making possible the identification and profiling of sRNAs in the cell The constantly increasing number of reads facilitates the characteri sation of the different pathways and reveals new classes of sRNAs In or der to process the large amount of sequences obtained from high through put experiments toolkits such as the UEA sRNA toolkit 21 were devel oped The UEA sRNA toolkit is available as a hosted service at srna tools cmp uea ac uk However as high throughput sequencing devices evolve and more reads are produced submitting the sequence data across the internet is becoming more problematic The UEA sRNA toolkit limits the input file size to manage UEA server resources and network bandwidth To overcome these restrictions the UEA bioinformatics group has packaged an open source stand alone version of the toolkit Users can download the toolkit and run the tools locally on Unix based desktops or servers mitigating the data transfer limitations to a hosted service This document describes the tools discusses their purpose and provides details on how to use them Audience This document and the toolkit are intended for bioinformaticians who should e Have a working knowledge of Linux Unix and running tools from the command line e Be familiar with small RNAs and their subtypes e Have basic knowledge of high throughput sequencing devices and the kind of out
49. put they produce particularly FASTQ and FASTA format files Document Organisation This user guide is organised as follows e Chapter 1 Installation contains details on what is provided by the UEA what the system pre requisities are and how to install the toolkit e Chapter 2 Quickstart contains a brief description of the tools and details how to run the tools from the command line e Chapter 3 Tools contains a detailed description of each tool in the sRNA toolkit e Chapter 4 Troubleshooting and FAQ contains answers to several commonly encountered error messages issues and ways to resolve them Notational Convention This user guide uses typefaces to identify the characteristics of text The general purpose typefaces and characteristics they imply are described in this table e Monospaced used for paths filenames commands and source code used for urls When describing tool parameters the typefaces are as follows e path parameters specify the location of a file or directory e g local usr myself tools e string e g tool adaptor adaptor_sequence_3 AGCTGGCTTC e numeric integer parameters have a default value and a range of allowed values If the input value is outside the range the default value is automatically assigned e g minsize 3 e numeric real parameters have a default value and a range of allowed values If the input value is outside the range the default va
50. se database to use miR Base databases can be found in INSTALL_PATH data directory The following files should be available File Name Description mature_all fa all mature miRNA sequences mature_animal fa mature miRNA sequences in Metazoa mature_plant fa mature miRNA sequences in Plants for each file an plusX variant is created that contains the ma ture sequences surrounded by XX at either end This allows the user to match with overhangs Only same strand matches of sRNAs to the miRBase databases will be reported miRBase databases can be downloaded and configured using the following command srna tools pl update_mirbase out The path to the output directory srna_file The location of the sRNA file in FASTA format e Optional collapse match groups Combines sRNAs and their counts based on their match signature The match signature of a sRNA is formed by combining all matching miRNA IDs i e a sRNA matching both miR156 and miR157 would have a match signature miR156 miR157 Each sRNA can be unambiguously assigned to one match signature genome If a genome FASTA file is provided sRNAs that do not have a genomic match are removed from the analysis group family If defined the matches to different members of the same family are combined into one group_mismatches If defined the matches to the same miRNA are combined into groups regardless of the number of mismatches e g
51. suite for R 29 Both files should contain the same time points treatments etc for the sRNA and gene data to be comparable The use of different points does not raise an error but the number of points must be identical Also the order of the constituent points is important because the next steps such as differential expression analysis and the correlation analysis are based on differences between consecutive points Please see our example files of 150 sRNA loci and genes in an experiment with 10 time points as a template for your own input files They can be found in INSTALL_PATH data and are called firepat_test150_genes csv and firepat_test150_srna_loci csv 3 6 2 Differential Expression Analysis FiRePat calculates correlation only for series that exhibit differential expres sion The top 1 of differentially expressed products are picked for further analysis where x is an input parameter de_threshold Set this param eter higher to include more profiles at the cost of a reduction in clustering accuracy Correlation and Clustering Pairs are created from highly co or anti regulated series from the two datasets SRNAs sRNA loci and genes For each pair the Pearson Corre lation Coefficient is computed If the degree of absolute similarity absolute value of the positive or negative correlation is above a given threshold sim_threshold the pair is selected for further analysis In order to fil ter and keep for further an
52. symbols as letters in the sequence A minimal FASTQ file might look like this SEQ_ID GATTTGGGGTTCAAAGCAGTATCGATCAAATAGTAAATCCATTTGTTCAACTCACAGTTT LR CCA LLL CLL Leek 2 55CCF gt gt gt gt gt gt CCCCCCCE65 What is a FASTA file In these tools we use the FASTA format for sRNA input files genomes and transcriptomes For short reads a FASTA file contains two lines for each read The first line shows the header of the read and the second the nucleotide sequence of the read An example file is shown below 46 gt ILLUMINA_READ_1 GGCCATCGAATATTA gt ILLUMINA_READ_2 GGTTTATGACACCTA The genomes and transcriptomes may use multiple lines for the nu cleotide sequence A FASTA file in redundant format is as follows gt ILLUMINA_READ_1 GGCCATCGAATATTA gt ILLUMINA_READ_1 GGCCATCGAATATTA gt ILLUMINA_READ_1 GGCCATCGAATATTA gt ILLUMINA_READ_1 GGCCATCGAATATTA gt ILLUMINA_READ_1 GGCCATCGAATATTA In a redundant file the abundance is not specified i e if a sequence is present than it was sequenced during the experiment A FASTA file in non redundant format represents the sequence abundance at the end of the sequence header gt identifier abundance sequence The example above in non redundant form is shown below gt ILLUMINA_READ_1 5 GGCCATCGAATATTA What is a CSV file The comma separated values CSV file format is a set of file formats used to store tabular data in which nu
53. t files significantly trrna If defined sRNAs matching sequences in the FASTA t r RNA file data t_and_r_RNAs fa will be removed uniq If defined adds columns to the output table for the number of reads in each locus and from each sample that had only a single hit to the reference genome This count can be used to filter loci and keep those with only unique matching sRNAs Example srna tools pl tool siloco genome data arabidopsis fa out output si sample_namel S1 sample_name2 S2 srna_filei files GSM118373_Rajagopalan_leaf fa srna_file2 files GSM154370_Carrington_col0_leaf fa asrp_links max_gap 100 maxsize 26 min_hits 5 minsize 20 num_samples 2 pseudocount 0 2 tair_links trrna uniq 33 The results are presented in a single csv file The header of the document contains the description of the data and read counts for samplel S1 and sample2 S2 The number of non redundant and redundant reads are listed for the input dataset and after each filtering step if any Valid sequences are those that passed the filter for size range low complexity t rRNA and genome matching The number of total valid reads is used for normalisation Locus data is shown in a table with the following columns e Chromosome start end position and length Genomic location and length of locus in nucleotides Some incomplete genomes may not yet be assembled into chromosomes and the acces sions listed here may be scaffolds or b
54. to the queue store genome data and third party binaries 60 All paths are relative to the doc root 61 i e the srna tools directory C2 E EEEEEEEEEEEEEEEEEEEEEEE EEE EEEE EEEE EEEE EEEE EEEE path to srna tools root directory on the cluster backend and the server The server root could be determined dynamically at run time but with this set up we can share resources between web apps by pointing them to the same root dir 69 It also creates a more unified interface to path data Ah e te oe te this is where jobs are send to get queued not the job storage area on the server 76 job _dir server We give a value for environmet server just for instant jobs where storage execution dir queue_job dir gt 81 server gt jobs 82 Figure 1 1 Configuring the config file application conf Chapter 2 Quickstart 2 1 Summary In this document the tools are referred to and described in the following order Low Level Tools applied on raw data e g FASTA sequences 1 2 Sequence file pre processing tool by S Moxon Converts read files from FASTQ to FASTA format and removes adap tor sequences making the input file ready for use by other tools Filter tool by F Schwach Filters sRNA sequence files in FASTA format according to user de fined criteria e g genome mapping reads specific size class t rRNA mapping reads miRCat tool miRNA Categoriser by S Mox
55. ur best to prepare such files Error reading input FASTA file The input FASTA file must be in redundant format as miRProf will count the occurrence of each sequence 50 4 2 5 RNA Hairpin Folding and Annotation Tool Pasting the example command gave error please correct param eter input Check line breaks on the command line A new line should only be used in the nucleotide data to separate the header from the sequence data The FASTA data must be enclosed in single quotes The following example shows where the line breaks should be srna tools pl tool hp_tool longSeq gt hairpin GGGAGCGGGGCTTCGATGATCGCTCGGTTTGAACGGATAGAGCGAATTCTGAGTGGTGCTCCC shortSeqs gt mirna GATAGAGCGAATTCTGAGTGGT out output h ol 4 2 6 FiRePat Short RNA file not recognised Unlike other tools in the tookit you must provide a csv table of expression values sRNA sequence Samplel Sample 2 Sample3 Sample4 AAAGTCGTA 10 l 100 25 Where can I find gene expression data for similar samples We have downloaded our example data both sRNA and gene data from GEO http www ncbi nlm nih gov geo 3 Why does it take so long run FiRePat creates all possible correlated pairs If the correlation threshold was low e g 90 then the number of pairs is large and the clustering step takes more time to complete How do I interpret the results The results are clustered on expression levels and the correlation
56. use not all sequencing devices include it in the resulting reads For example for 454 datasets and conventional cloning and capillary sequencing both the 5 and the 3 adaptors are in cluded in the input file In contrast Solexa Tllumina reads start at the first base of the sRNA and contain only the 3 adaptor see figure 3 1 The tool only looks for exact matches to the adaptor sequence s so it will not remove adaptors containing mismatches For this reason it is often prefer able to provide a truncated version of the adaptor sequence as input For example the first 8nt of the adaptor sequence are sufficient for 3 adaptor matching or the last 8nt of the adaptor sequence are sufficient for 5 adaptor matching Parameters e Required adaptor_sequence_3 The 3 adaptor sequence srna file The location of the sRNA file in FASTQ or FASTA for mat 15 Solexa read 5 adaptor SEQUENCE 3 adaptor Figure 3 1 Read with adaptors A Solexa Tllumina read starts at the first base of the sRNA and contains only the first part of the 3 adaptor out The path to the output directory e Optional adaptor_sequence_5 The 5 adaptor sequence allow_rev_comp If used matches to the reverse complement of adaptor sequences are allowed This parameter is only required for classical capillary sequencing where the orientation of the clone relative to the sequencing primers is not known minsize The mi
57. vels for each sequence miRProf generates two files a results table in csv format and a list of sRNAs in FASTA format that match known miRNAs The results table contains a formatted list of reads that match to known miRNAs It also contains information about redundant total and non redundant unique sequence counts in the input set before and after every filtering step The total abundance of reads after the final filtering step is used for normalisation 8 Normalised counts are given in matching reads per 1 million total reads RPM to make them comparable between samples The rest of the table lists miRNA matches and associated sequence counts Small RNAs with matches to multiple miRNAs or miRNA hairpins receive a weighted match count that is obtained by dividing the raw count by the number of matches The FASTA file contains the actual sequences from your file The ID lines contain the following information gt mirnaIDs_n_c miRNA_sequence where mirnalDs is the identifier obtained by concatenating the IDs of matching miRNAs n is the consecutive number for each match and c is the raw count for the matching sequence Expression profiles of reads can be produced by running miRProf separately on mutliple samples and merging the results tables sRNA sequence Samplel Sample 2 Sample3 Sample4 miRNA_1 10 1 100 25 miRNA_2 100 10 20 55 Parameters 24 e Required mirbase_db The location of the miRBa
58. with expression levels N A SiLoCo 2x FASTA sRNA file 1x FASTA Genome SiLoMa 1x FASTA sRNA file 1x FASTA Genome ta siRNA prediction 1x FASTA sRNA file 1x FASTA Genome Plant target prediction 1x FASTA sRNA file 1x FASTA Transcriptome Given the size of the genomes these are not distributed with the toolkit Frequently used genomes can be downloaded from the following URLs e Arabidopsis Thaliana ftp ftp arabidopsis org home tair Genes TAIR9_genome_release TAIR9_chr_All fas e Solanum Lycopersicum http solgenomics net genomes Solanum_lycopersicum index P e Oryza Sativa http rice genomics org cn rice link download jsp o uman Pp o ouse http hgdownload cse ucsc edu downloads html 10 ttp hgdownload cse ucsc edu downloads html e Drosophila http www fruitfly org sequence download html The running times for each of the tools on a Dual Quad core Intel Xeon 2 50GHz L5420 with 32GB RAM Linux server are Tool Plant Animal Runtime mins Memory Mb Runtime mins Memory Gb Pre processing 5 64 10 64 Filter 7 64 14 64 miRCat 125 256 150 256 miRProf T 64 11 64 RNA folding instant 64 instant 64 FiRePat instant 64 instant 64 SiLoCo 9 64 15 64 SiLoMa 8 64 11 64 ta siRNA prediction 11 64 N A N A Plant target prediction gt 4days 1000 N A N A Note The run times were computed using the default values for
Download Pdf Manuals
Related Search