Bismark Bisulfite Mapper – User Guide - v0.7.7
Contents
1. The Bismark methylation extractor comes with a few options such as ignoring the first lt int gt number of positions in the methylation call string e g to remove a restriction enzyme site if RRBS is performed with non directional BS Seq libraries it might be required to remove reconstituted Mspl sites at the beginning of each read as they will introduce a bias into the first methylation call please send me an email if you require further information on this Another useful option for paired end reads is called no overlap specifying this option will extract the methylation calls of overlapping parts in the middle of paired end reads only once using the calls from the first read which is presumably the one with a lowest error rate For a full list of options type methylation extractor help at the command line or refer to the Appendix section at the end of this User Guide Methylation extractor output By default the Bismark methylation extractor discriminates between cytosines in CpG CHG or CHH context If desired CHG and CHH contexts can be merged into a single non CpG context by specifying the option merge non CpG as a word of warning this might produce files with up to several hundred million lines Strand specific methylation output files default As its default option the methylation extractor will produce a strand specific output which will use the following abbreviations in the output file name to2. vanilla Bismark output Bowtie 1 only Bismark can generate a comprehensive alignment and methylation call output file for each input file or set of paired end input files The sequence basecall qualities of the input FastQ files are written out into the Bismark output file as well to allow filtering on quality thresholds Please note that the quality values are encoded in Sanger format Phred 33 scale even if the input was in Phred64 or the old Solexa format The single end output contains the following information 1 line per sequence tab separated seq ID alignment strand chromosome start end original bisulfite read sequenc equivalent genomic sequence 2 extra bp methylation call string read conversion genome conversion eea ce co na E O G 00 u DOB WN FP Dat i e a veo reo YY HY read quality score Phred33 scale Single end alignment example HWUSI EAS611 0001 3 1 1186 18334 0 1 4 122893213 122893242 TGGGTTTTTAAGATTTTGTGTAGTTGGGGTTTGGAGATGG CGGGCCCTCAAGACCCTGCGCAGCTGGGGCCTGGAGACGGAG Zes WON Hees DNR a2 Ke Ke eres lob narra Zone CT CI ITIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII H O g OTA ABP WYN FP e pee aaaaa aana rye vrvrvrvr vr YH YH YH YS YH The paired end output looks like this 1 line per sequence pair tab separated 1 seq ID 2 alignment strand 3 chromosome 4 start 5 end 6 original bisulfite read sequence 1 7 equivalent genom
3. but that does not satisfy the paired end constraints fr rf ff I X This option disables that behaviour and is on by default Up to lt int gt consecutive seed extension attempts can fail before Bowtie 2 moves on using the alignments found so far A seed extension fails if it does not yield a new best or a new second best alignment Default 15 lt int gt is the maximum number of times Bowtie 2 will re seed reads with repetitive seeds When re seeding Bowtie 2 simply chooses a new set of reads same length same number of mismatches allowed at different offsets and searches for more alignments A read is considered to have repetitive seeds if the total number of seed hits divided by the number of seeds that aligned at least once is greater than 300 Default 2 Bowtie 2 parallelization options p NTHREADS Launch NTHREADS parallel search threads default 1 Threads will run on separate processors cores and synchronize when parsing reads and outputting alignments Searching for alignments is highly parallel and speedup is close to linear NOTE It is currently unclear whether this speed increase also translates into a speed increase of Bismark since it is running several instances of Bowtie 2 concurrently Increasing p increases Bowtie 2 s memory footprint E g when aligning to a human genome index increasing p from 1 to 8 increases the memory footprint by a few hundred megabytes for each instance of Bowti
4. for Bowtie 2 see L Maximum permitted total of quality values at all mismatched read positions throughout the entire alignment not just in the seed The default is 70 Like Maq Bowtie rounds quality values to the nearest 10 and saturates at 30 The number of megabytes of memory a given thread is given to store path descriptors in best mode Best first search must keep track of many paths at once to ensure it is always extending the path with the lowest cumulative cost Bowtie tries to minimize the memory impact of the descriptors but they can still grow very large in some cases If you receive an error message saying that chunk memory has been exhausted in best mode try adjusting this parameter up to dedicate more memory to the descriptors Default 512 The minimum insert size for valid paired end alignments E g if I 60 is specified and a paired end alignment consists of two 20 bp alignments in the appropriate orientation with a 20 bp gap between them that alignment is considered valid as long as X is also satisfied A 19 bp gap would not be valid in that case Default 0 The maximum insert size for valid paired end alignments E g if X 100 is specified and a paired end alignment consists of two 20 bp alignments in Bowtie 1 Reporting k lt 2 gt best no_ best Output the proper orientation with a 60 bp gap between them that alignment is considered valid as long as I is also satisfied A61 bp gap w
5. methylation percentage is reported Default 1 remove_ spaces Replaces whitespaces in the sequence ID field with underscores to allow sorting counts Adds two additional columns to the output file to enable further calculations col 5 number of methylated calls col 6 number of unmethylated calls This option is required if cytosine report is specified and will be set automatically if necessary CX CX_ context The sorted bedGraph output file contains information on every single cytosine that was covered in the experiment irrespective of its sequence context This applies to both forward and reverse strands Please be aware that this option may generate large temporary and output files and may take a long time to sort up to many hours Default OFF i e Default CpG context only Genome wide cytosine methylation report specific options cytosine report After the conversion to bedGraph has completed the option cytosine report produces a genome wide methylation report for all cytosines in the genome By default the output uses 1 based chromosome coordinates zero based cords are optional and reports CpG context only all cytosine context is optional The output considers all Cs on both forward and reverse strands and reports their position strand trinucleotide content and methylation state counts are 0 if not covered CX CX context The output file contains information on every single cytosine in the genome ir
6. mode or eight strand specific output files default for Cs in i CpG context ii any non CpG context Depending on the C content of the Bismark result file the output file size might reach 10 30GB no header Suppresses the Bismark version header line in all output files for more convenient batch processing o output DIR Allows specification of a different output directory absolute or relative path If not specified explicitely the output will be written to the current directory report Prints out a short methylation summary and the parameters used to run this script version Displays the version information h help Displays this help file and exits bedGraph specific options bedGraph After finishing the methylation extraction the methylation output is written into a sorted bedGraph file that reports the position of a given cytosine and its methylation state in seem details below The methylation extractor output is temporarily split up into temporary files one per chromosome written into the current directory or folder specified with o output these temp files are then used for sorting and deleted afterwards By default only cytosines in CpG context will be sorted The option CX_ context may be used to report all cyosines irrespective of sequence context this will take MUCH longer cutoff threshold The minimum number of times a methylation state has to be seen for that nucleotide before its
7. 0 data genomes homo_sapiens GRCh37 s_ 1 sequence txt What does the Bismark output look like As of version 0 6 x the default output of Bismark is in SAM format when using either Bowtie 1 or Bowtie 2 The former custom Bismark output for Bowtie 1 which used to be the standard output up to versions 0 5 x is still available by specifying the option vanilla see below The Bismark output using Bowtie 2 is invariably in SAM format required to encode gapped alignments Bismark SAM output default By default Bismark generates SAM output for all alignment modes Please note that reported quality values are encoded in Sanger format Phred 33 scale even if the input was in Phred64 or the old Solexa format 1 QNAME segq ID 2 FLAG this flag tries to take the strand a bisulfite read originated from into account this is different from ordinary DNA alignment flags 3 RNAME chromosome 4 POS start position 5 MAPQ always 255 6 CIGAR 7 RNEXT 8 PNEXT 9 TLEN 0 SEQ 1 QUAL Phred33 scale 2 NM tag edit distance to the reference 3 XX tag base by base mismatches to the reference not including 4 XM tag methylation call string 5 XR tag read conversion state for the alignment 6 XG tag genome conversion state for the alignment The mate read of paired end alignments is written out as an additional separate line in the same format Custom
8. 1 and Bowtie 2 require distinct indexing steps since their indexes are not compatible II Bismark read alignment step Simply specify a file to be analysed a reference genome and alignment parameters Bismark will produce a combined alignment methylation call output default is SAM format as well as a run statistics report III Bismark methylation extractor This step is optional and will extract the methylation information from the Bismark alignment output Running this additional step allows splitting the methylation information up into the different contexts getting strand specific methylation information and offers some filtering options Each of these steps will be described in more detail with examples in the following sections I Bismark Genome Preparation This script needs to be run only once to prepare the genome of interest for bisulfite alignments You need to specify a directory containing the genome you want to align your reads against please be aware that the bismark genome preparation script currently expects FastaA files in this folder with either fa or fasta extension single or multiple sequence entries per file Bismark will create two individual folders within this directory one for a C gt T converted genome and the other one for the G gt A converted genome After creating C gt T and G gt A versions of the genome they will be indexed in parallel using the indexer bowtie build or bowtie2 build Once both C
9. OB strands thereby ignoring alignments coming from the complementary strands as they should theoretically not be present in the BS Seq library in question b Alternatively BS Seq libraries can be constructed so that all four different strands generated in the BS PCR can and will end up in the sequencing library with roughly the same likelihood In this case all four strands OT CTOT OB CTOB can produce valid alignments and the library is called non directional Specifying non directional instructs Bismark to use all four alignment outputs To summarise this again alignments to the original top strand or to the strand complementary to the original top strand OT and CTOT will both yield methylation information for cytosines on the top strand Alignments to the original bottom strand or to the strand complementary to the original bottom strand OB and CTOB will both yield methylation information for cytosines on the bottom strand i e they will appear to yield methylation information for G positions on the top strand of the reference genome For more information about how to extract methylation information of the four different alignment strands please see below in the section on the Bismark methylation extractor USAGE bismark options lt genome folder gt 1 lt matesl1 gt 2 lt mates2 gt lt singles gt A typical single end analysis of a 40 bp sequencing run could look like this bismark q phred64 quals n 1 1 4
10. Oct 02 2012 A d Bismark Bisulfite Mapper User Guide v0 7 7 1 Quick Reference Bismark needs a working version of Perl and it is run from the command line Furthermore Bowtie http bowtie bio sourceforge net index shtml or Bowtie 2 http bowtie bio sourceforge net bowtie2 needs to be installed on your computer For more information on how to run Bismark with Bowtie 2 please go to the end of this manual First you need to download a reference genome and place it in a genome folder Genomes can be obtained e g from the Ensembl http www ensembl org info data ftp index html or NCBI websites ftp ftp ncbi nih gov genomes for the example below you would need to download the Homo sapiens genome Bismark supports reference genome sequence files in FastA format allowed file extensions are either either fa or fasta Both single entry and multiple entry FastaA files are supported The following examples will use the file test dataset fastq which is available for download from the Bismark homepage it contains 10 000 reads in FastQ format Phred33 qualities 50 bp long reads from a human directional BS Seq library I Running bismark_genome_preparation USAGE bismark genome preparation options lt path_to genome folder gt A typical genome indexing could look like this bismark bismark genome preparation path to bowtie usr local bowtie verbose data genomes homo_sapiens GRCh37 Il Running bis
11. ads will appear as they did in the input without any of the translation of quality values that may have taken place within Bowtie or Bismark Paired end reads will be written to two parallel files with _1 and _2 inserted in their filenames i e ambiguous reads 1 txt and _ambiguous_ reads 2 txt fq These reads are not written to the file specified with un o output_dir lt dir gt Write all output files into this directory By default the output files will be temp dir lt dir gt Other h help v version written into the same folder as the input file If the specified folder does not exist Bismark will attempt to create it first The path to the output folder can be either relative or absolute Write temporary files to this directory instead of into the same directory as the input files If the specified folder does not exist Bismark will attempt to create it first The path to the temporary folder can be either relative or absolute Displays this help file Displays version information BOWTIE 2 SPECIFIC OPTIONS bowtie2 Uses Bowtie 2 instead of Bowtie 1 Bismark limits Bowtie 2 to only perform end to end alignments i e searches for alignments involving all read characters also called untrimmed or unclipped alignments Bismark assumes that raw sequence data is adapter and or quality trimmed where appropriate Default off Bowtie 2 alignment options N lt int gt L lt int gt Sets the num
12. ber of mismatches to be allowed in a seed alignment during multiseed alignment Can be set to 0 or 1 Setting this higher makes alignment slower often much slower but increases sensitivity Default 0 This option is only available for Bowtie 2 forBowtie 1 see n Sets the length of the seed substrings to align during multiseed alignment Smaller values make alignment slower but more senstive Default the sensitive preset of Bowtie 2 is used by default which sets L to 20 This option is only available for Bowtie 2 for Bowtie 1 see 1 ignore quals When calculating a mismatch penalty always consider the quality value at the mismatched position to be the highest possible regardless of the actual value l e input is treated as though all quality values are high This is also the default behaviour when the input doesn t specify quality values e g in mode For bisulfite alignments in Bismark this option is invariable and on by default Bowtie 2 paired end options no mixed no discordant Bowtie 2 Effort options D lt int gt R lt int gt This option disables Bowtie 2 s behaviour to try to find alignments for the individual mates if it cannot find a concordant or discordant alignment for a pair This option is invariable and on by default Normally Bowtie 2 looks for discordant alignments if it cannot find any concordant alignments A discordant alignment is an alignment where both mates align uniquely
13. e 2 This option is only available if Bowtie is linked with the pthreads library i e if BOWTIE_PTHREADS 0 is not specified at build time In addition this option will automatically use the option reorder which guarantees that output SAM records are printed in an order corresponding to the order of the reads in the original input file even when p is set greater than 1 Bismark requires the Bowtie 2 output to be this way Specifying reorder and setting p greater than 1 causes Bowtie 2 to run somewhat slower and use somewhat more memory then if reorder were not specified Has no effect if p is set to 1 since output order will naturally correspond to input order in that case Bowtie 2 Scoring options score min lt func gt Sets a function governing the minimum alignment score needed for an alignment to be considered valid i e good enough to report This is a function of read length For instance specifying L 0 0 2 sets the minimum score function f to f x 0 0 2 x where x is the read length See also setting function options at http bowtie bio sourceforge net bowtie2 The default is L 0 0 2 Bowtie 2 Reporting options most valid alignments lt int gt This used to be the Bowtie 2 parameter M As of Bowtie 2 version 2 0 0 beta7 the option M is deprecated It will be removed in subsequent versions What used to be called M mode is still the default mode but adjusting the M setting is deprecated Use
14. e g flyA_1 fq flyB_1 fq Sequences specified with this option must correspond file for file and read for read with those specified in lt mates2 gt Reads may bea mix of different lengths Bismark will produce one mapping result and one report file per paired end input file pair Comma separated list of files containing the 2 mates filename usually includes _2 e g flyA_1 fq flyB_1 fq Sequences specified with this option must correspond file for file and read for read with those specified in lt mates1 gt Reads may be a mix of different lengths A comma or space separated list of files containing the reads to be aligned e g lanel fq lane2 fq lane3 f q Reads may be a mix of different lengths Bismark will produce one mapping result and one report file per input file The query input files specified as lt mate1 gt lt mate2 gt or lt singles gt are FastQ files usually having extension fqor fasta This is the default See also solexa quals and integer quals The query input files specified as lt mate1 gt lt mate2 gt or lt singles gt are FastA files usually having extension fa mfa fna or similar All quality values are assumed to be 40 on the Phred scale Skip i e do not align the first lt int gt reads or read pairs from the input Only aligns the first lt int gt reads or read pairs from the input Default no limit phred33 quals phred64 quals solexa quals solexal 3 qual
15. ethylation_extractor USAGE methylation extractor options lt filenames gt A typical command for a single end file could look like this methylation extractor s comprehensive s_1 sequence txt_bismark sam If you get stuck at any point or have any questions or comments please contact me via email felix krueger babraham ac uk 4 APPENDIX Full list of options I Bismark Genome Preparation A full list of options can also be viewed by typing bismark genome preparation help USAGE bismark genome preparation options lt arguments gt OPTIONS help man verbose yes yes to all path_ to bowtie lt bowtie2 single fasta ARGUMENTS lt path_to_ genome folder gt Displays this help file Print verbose output for more details or debugging Answer yes to safety related questions such as Are you sure you want to overwrite any existing files in the Bisulfite_Genomes folder gt The full path to your Bowtie 1 Bowtie 2 installation If the path is not provided as an option you will be prompted for it later This will create bisulfite indexes for Bowtie 2 Default Bowtie 1 Instruct the Bowtie Indexer to write the converted genomes into single entry FastA files instead of making one multi FastA file MFA per chromosome This might be useful if individual bisulfite converted chromosomes are needed e g for debugging however it can cause a problem with indexing if the n
16. gt T and G gt A genome indices have been created you do not need to use the genome preparation script again unless you want to align against a different genome Please note that Bowtie 1 and 2 indexes are not compatible To create a genome index for use with Bowtie 2 the option bowtie2 needs to be included as well Running bismark_genome_preparation USAGE bismark genome preparation options lt path_to genome folder gt A typical command could look like this bismark genome preparation path to bowtie usr local bowtie verbose data genomes homo_sapiens GRCh37 II Bismark alignment step This step represents the actual bisulfite alignment and methylation calling part Bismark requires the user to specify only two things The directory containing the genome of interest This folder must contain the unmodified genome as fa or fasta files as well as the two bisulfite genome subdirectories which were generated in the Bismark Genome Preparations step see above The sequence file s to be analysed in either FastQ or FastA format All other parameters are optional In the current version it is required that the current working directory also contains the sequence files to be analysed For each sequence file or each set of paired end sequence files Bismark produces one alignment and methylation call output file as well as a report file detailing alignment and methylation call statistics for your information and reco
17. he read The files may be gzip compressed ending in gz Input file s are Bismark result file s generated from single end read data Specifying either single end or paired end is mandatory Input file s are Bismark result file s generated from paired end read data Specifying either paired end or single end is mandatory The Bismark result input file s are in the old custom Bismark format up to version 0 5 x and not in SAM format which is the default as of Bismark version 0 6 x or higher Default OFF For paired end reads it is theoretically possible that read_1 and read_2 overlap This option avoids scoring overlapping methylation calls twice Whilst this removes a bias towards more methylation calls towards the center of sequenced fragments it can de facto remove a good proportion of the data Ignore the first lt int gt bp when processing the methylation call string As all reads are sorted in a forward direction this can remove e g a restriction enzyme site at the start of each read this can become relevant for non directional libraries in RRBS type experiments Specifying this option will merge all four possible strand specific methylation info into context dependent output files The default contexts are i CpG context ii CHG context iii CHH context Depending on the C content of the Bismark result file the output file size might reach 10 30GB This will produce two output files in comprehensive
18. ic sequence 1 2 extra bp 8 methylation call string 1 9 original bisulfite read sequence 2 10 equivalent genomic sequence 2 2 extra bp 11 methylation call string 2 12 read 1 conversion 13 genome conversion 14 read 1 quality score Phred33 scale 15 read 2 quality score Phred33 scale Paired end alignment example HWUSI EBAS611 100205 2 1 13 1732 0 4 14 62880539 62880652 CGGGATTTCGCGGAGTACGGGTGATCGTGTGGAATATAGA CGGGACTCCGCGGAGCACGGGTGACCGTGTGGAATACAGAGT Lice eae Ne Ke Lise SA le ies BL laa ite wt oe Kulsa CAACTATCTAAAACTAAAAATAACGCCGCCCAAAAACTCT TCCGGCTGTCTGGAGCTGAAGATGGCGCCGCCCAGAAGCTCT sO sie eee ok AE is Ae E EAE S EE CL CT ITIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII PrP FP PrP Pee ee RO OOO Oe wW U DOO WAAIA TF WN FB rYyeorovrvrvrvrvr vr reo vrorvrwr wH m we eee Sey een eS Data visualisation To see the location of the mapped reads the Bismark output file can be imported into a genome viewer such as SeqMonk using the chromosome start and end positions this can be useful to identify regions in the genome which display an artefactually high number of aligned reads The alignment output can also be used to apply post processing steps such as de duplication allowing only 1 read for each position in the genome to remove PCR artefacts or filtering on the number of bisulfite conversion related non bisulfite mismatches please note that s
19. indicate the strand the alignment came from OT original top strand CTOT complementary to original top strand OB original bottom strand CTOB complementary to original bottom strand Methylation calls from OT and CTOT will be informative for cytosine methylation positions on the original top strand calls from OB and CTOB will be informative for cytosine methylation positions on the original bottom strand Please note that specifying the directional option in the Bismark alignment step will not report any alignments to the CTOT or CTOB strands As cytosines can exist in any of three different sequence contexts CpG CHG or CHH the methylation extractor default output will consist of 12 individual output files per input file CoG OT 25 CPG CTOT sasur CPG OB s Ste Context dependent methylation output files comprehensive option If strand specific methylation is not of interest all available methylation information can be pooled into a single context dependent file information from any of the four strands will be pooled This will default to three output files CpG context CHG context and CHH context or result in 2 output files CpG context and Non CpG context if merge non CpG was selected note that this can result in enormous file sizes for the non CpG output Both strand specific and context dependent options can be combined with the merge non CpG option II Running the Bismark m
20. lignment score as reported by the Bowtie 2 AS i field If a read produces several alignments with the same number of mismatches or with the same alignment score AS i field a read or a read pair is discarded altogether Bismark alignment and methylation call report Upon completion Bismark produces a run report containing information about the following Summary of alignment parameters used Number of sequences analysed Number of sequences with a unique best alignment mapping efficiency Statistics summarising the bisulfite strand the unique best alignments came from Number of cytosines analysed Number of methylated and unmethylated cytosines Percentage methylation of cytosines in CpG CHG or CHH context where H can be either A T or C This percentage is calculated individually for each context following the equation methylation context 100 methylated Cs context methylated Cs context unmethylated Cs context It should be stressed that the percent methylation value context is just a very rough calculation performed directly at the mapping step Actual methylation levels after post processing or filtering have been applied may vary 3 Running Bismark Running Bismark is split up into three individual steps I First the genome of interest needs to be bisulfite converted and indexed to allow Bowtie alignments This step needs to be carried out only once for each genome Note that Bowtie
21. mark USAGE bismark options lt genome folder gt 1 lt matesl gt 2 lt mates2 gt lt singles gt Typical alignment example tolerating one non bisulfite mismatch per read bismark n 1 1 50 data genomes homo_ sapiens GRCh37 test dataset fastq This will produce two output files a test dataset fastq bismark sam contains all alignments plus methylation call strings b test dataset fastq bismark mapping report txt contains alignment and methylation summary NOTE In order to work properly the current working directory must contain the sequence files to be analysed IlI Running the Bismark methylation_extractor USAGE methylation_extractor options lt filenames gt A typical command to extract context dependent CpG CHG CHH methylation could look like this methylation extractor s comprehensive test dataset fastq bismark sam This will produce three output files a CpG context test dataset fastq bismark sam txt b CHG context test dataset fastq bismark sam txt c CHH context test dataset fastq bismark sam txt 2 Bismark General Information What is Bismark Bismark is a set of tools for the time efficient analysis of Bisulfite Seq BS Seq data Bismark performs alignments of bisulfite treated reads to a reference genome and cytosine methylation calls at the same time Bismark is written in Perl and is run from the command line Bisulfite treated reads are mapped using the short read aligner B
22. ntext gt lt trinucleotide context gt
23. o the original strands are merely theoretical and should not exist in reality Specifying directional alignments which is the default will only run 2 alignment threads to the original top OT or bottom OB strands in parallel and report these alignments This is the recommended option for sprand specific libraries Suppress SAM header lines starting with This might be useful when very large input files are split up into several smaller files to run concurrently and the output files are to be merged afterwards Print nothing besides alignments Performs bisulfite mapping with Bowtie 1 and prints the old custom Bismark output up to versions 0 5 X instead of SAM format output Write all reads that could not be aligned to the file unmapped _reads txt in the output directory Written reads will appear as they did in the input without any translation of quality values that may have taken place within Bowtie or ambiguous Bismark Paired end reads will be written to two parallel files with _1 and _2 inserted in their filenames ie unmapped reads 1 txt and unmapped reads_2 txt Reads with more than one valid alignment with the same number of lowest mismatches ambiguous mapping are also written to unmapped reads txt unless ambiguous is also specified Write all reads which produce more than one valid alignment with the same number of lowest mismatches or other reads that fail to align uniquely to _ambiguous_reads txt Written re
24. on the size of the reference genome For a large eukaryotic genome human or mouse we experienced a typical memory usage of around 12GB We thus recommend running Bismark on a machine with 5 CPU cores and at least 12 GB of RAM The memory requirements of Bowtie 2 are somewhat larger possibly to allow gapped alignments When running Bismark using Bowtie 2 we therefore recommend a system with at least 5 cores and gt 16GB of RAM Alignment speed depends largely on the read length and alignment parameters used Allowing many mismatches and using a short seed length which is the default option for Bowtie 1 see below tends to be fairly slow whereas looking for near perfect matches can align around 5 25 million sequences per hour using Bowtie 1 Since we haven t tested Bowtie 2 very much yet we can t give recommendations about alignment parameters and speed of Bowtie 2 at the current time BS Seq test data set A test BS Seq data set is available for download from the Bismark homepage It contains 10 000 single end shotgun BS reads from human ES cells in FastQ format from SRRO20138 Lister et al 2009 trimmed to 50 bp base call qualities are Sanger encoded Phred values Phred33 Which kind of BS Seq files and or experiments are supported Bismark supports the alignment of bisulfite treated reads whole genome shotgun BS Seq WGSBS or reduced representation BS Seq RRBS for the following conditions sequence format either FastQ o
25. ould not be valid in that case Default 500 Due to the way Bismark works Bowtie 1 will report up to 2 valid alignments This option is used by default and cannot be changed Make Bowtie guarantee that reported singleton alignments are best in terms of stratum i e number of mismatches or mismatches in the seed in the case if n mode and in terms of the quality e g a 1 mismatch alignment where the mismatch position has Phred quality 40 is preferred over a 2 mismatch alignment where the mismatched positions both have Phred quality 10 When best isnot specified Bowtie may report alignments that are sub optimal in terms of stratum and or quality though an effort is made to report the best alignment best mode also removes all strand bias Note that best does not affect which alignments are considered valid by Bowtie only which valid alignments are reported by Bowtie Bowtie is about 1 2 5 times slower when best is specified Default on Disables the best option which is on by default This can speed up the alignment process e g for testing purposes but for credible results it is not recommended to disable best non directional The sequencing library was constructed in a non strand specific manner alignments sam no hd quiet vanilla to all four bisulfite strands will be reported Default OFF The current Illumina protocol for BS Seq is directional in which case the strands complementary t
26. out to a new output file depending on its context CpG CHG or CHH whereby methylated Cs will be labelled as forward reads non methylated Cs as reverse reads The resulting files can be imported into a genome viewer such as SeqMonk using the generic text import filter and the analysis of methylation data can commence Alternatively the output of the methylation extractor can be transformed into a bedGraph file using the script genome_methylation_bismark2bedGraph which was kindly provided by Dr Oliver Tam also available for download at www bioinformatics babraham ac uk projects bismark As of Bismark version 0 6 or higher the default input format for the methylation extractor is SAM The former custom Bismark format can still be used as well by specifying vanilla The methylation extractor output looks like this tab separated 1 seq ID 2 methylation state 3 chromosome 4 start position end position 5 methylation call Methylated cytosines will receive a orientation unmethylated cytosines will receive a orientation Examples for cytosines in CpG context HWUSI EAS611 0006 3 1 1058 15806 0 1 6 91793279 z HWUSI EAS611_ 0006 3 1 1058 17564 0 1 8 122855484 Z Examples for cytosines in CHG context HWUSI EAS611 0006 3 1 1054 1405 0 1 7 89920171 x HWUSI EAS611 0006 3 1 1054 1405 0 1 7 89920172 X Examples for cytosines in CHH context HWUSI EAS611_ 0006 3 1 1054 1405 0 1 7 89920184 h
27. owtie 1 Langmead B Trapnell C Pop M Salzberg SL Ultrafast and memory efficient alignment of short DNA sequences to the human genome Genome Biol 10 R25 or alternatively Bowtie 2 and therefore it is a requirement that Bowtie 1 or Bowtie 2 are also installed on your machine see Dependencies All files associated with Bismark as well as a test BS Seq data set can be downloaded from http www bioinformatics babraham ac uk projects bismark We would like to hear your comments suggestions or bugs about Bismark Please email them to felix krueger babraham ac uk Installation notes Bismark is written in Perl and is executed from the command line To install Bismark simply copy the bismark_vO X Y tar gz file into a Bismark installation folder and extract all files by typing tar xzf bismark vO X Y tar gz Dependencies Bismark requires a working of Perl and Bowtie 1 Bowtie 2 to be installed on your machine http bowtie bio sourceforge net index shtml http bowtie bio sourceforge net bowtie2 Bismark will assume that the Bowtie 1 Bowtie 2 executable is in your path unless the path to Bowtie is specified manually with path_to bowtie lt bowtie gt NOTE In order to work properly the current working directory must contain the sequence files to be analysed Hardware requirements Bismark holds the reference genome in memory and in addition to that runs four parallel instances of Bowtie The memory usage is dependent
28. r FastA single end or paired end reads input files can be uncompressed or gzip compressed ending in gz variable read length support directional or non directional BS Seq libraries In addition Bismark retains much of the flexibility of Bowtie 1 Bowtie 2 adjustable seed length number of mismatches insert size For a full list of options please run bismark help or see the Appendix at the end of this User Guide NOTE It should be mentioned that Bismark supports only reads in base space such as from the Illumina platform There are currently no plans to extend its functionality to color space reads How does Bismark work Sequence reads are first transformed into fully bisulfite converted forward C gt T and reverse read G gt A conversion of the forward strand versions before they are aligned to similarly converted versions of the genome also C gt T and G gt A converted Sequence reads that produce a unique best alignment from the four alignment processes against the bisulfite genomes which are running in parallel are then compared to the normal genomic sequence and the methylation state of all cytosine positions in the read is inferred For use with Bowtie 1 a read is considered to align uniquely if one alignment exists that has with fewer mismatches to the genome than any other alignment or if there is no other alignment For Bowtie 2 a read is considered to align uniquely if an alignment has a unique best a
29. rd keeping Running bismark Before running Bismark we recommend spending some time on quality control of the raw sequence files using FastQC www bioinformatics babraham ac uk projects fastqc Fastac might be able to spot irregularities associated with your BS Seq file such as high base calling error rates or contaminating sequences such as PCR primers or Illumina adapters Many sources of error impact detrimentally the alignment efficiencies and or alignment positions and thus possibly also affect the methylation calls and conclusions drawn from the experiment If no additional options are specified Bismark will use a set of default values some of which are Using Bowtie 1 If no specific path to Bowtie is specified it is assumed that the bowtie executable is in the PATH Bowtie 1 is run best mode it is possible but not recommended to turn this off Standard alignments allow up to 2 mismatches in the seed region which is defined as the first 28 bp by default These parameters can be modified using the options n and 1 respectively Using Bowtie 2 If no specific path to Bowtie 2 is specified it is assumed that the bowtie2 executable is in the PATH Standard alignments use a multi seed length of 20bp with O mismatches These parameters can be modified using the options L and N respectively Standard alignments report the best of up to 10 valid alignments This can be set using the M parameter Standard alignments u
30. respective of its context This applies to both forward and reverse strands Please be aware that this will generate output files with gt 1 1 billion lines for a mammalian genome such as human or mouse Default OFF i e Default CpG context only zero_ based Uses zero based coordinates like used in e g bed files instead of 1 based coordinates Default OFF genome folder lt path gt Enter the genome folder you wish to use to extract sequences from full path only Accepted formats are FastA files ending with fa or fasta Specifying a genome folder path is mandatory split by chromosome Writes the output into individual files for each chromosome instead of a single output file Files will be named to include the input filename and the chromosome number The bismark_methylation_extractor output is in the form tab delimited lt seq ID gt lt methylation state gt lt chromosome gt lt start position end position gt lt methylation call gt Methylated cytosines receive a orientation Unmethylated cytosines receive a orientation The bedGraph output optional looks like this tab delimited lt chromosome gt lt start position gt lt end position gt lt methylation percentage gt The genome wide cytosine methylation output file optional is tab delimited in the following format lt chromosome gt lt position gt lt strand gt lt count methylated gt lt count non methylated gt lt C co
31. s path_to bowtie Alignment FastQ qualities are ASCII chars equal to the Phred quality plus 33 Default on FastQ qualities are ASCII chars equal to the Phred quality plus 64 Default off Convert FastQ qualities from solexa scaled which can be negative to phred scaled which can t The formula for conversion is phred qual 10 log 1 10 solexa qual 10 0 log 10 Used with q This is usually the right option for use with unconverted reads emitted by the GA Pipeline versions prior to 1 3 Default off Same as phred64 qual1s This is usually the right option for use with unconverted reads emitted by GA Pipeline version 1 3 or later Default off The full path lt gt tothe Bowtie 1 or 2 installation on your system If not specified it will be assumed that Bowtie is in the path n seedmms lt int gt The maximum number of mismatches permitted in the seed i e the first 1 seedlen e magerr lt int gt chunkmbs lt int gt I minins lt int gt X maxins lt int gt base pairs of the read where Lis set with 1 seedlen This may be 0 1 2 or 3 and the default is 2 This option is only available for Bowtie 1 for Bowtie 2 see N The seed length i e the number of bases of the high quality end of the read to which the n ceiling applies The default is 28 Bowtie and thus Bismark is faster for larger values of 1 This option is only available for Bowtie 1
32. se the default minimum alignment score function L 0 0 6 i e f x 0 0 6 x where x is the read length For a read of 75bp this would mean that a read can have a lowest alignment score of 45 before an alignment would become invalid This is roughly equal to 7 mismatches or 4 indels of 3 4 bp in the read or a combination thereof The stringency can be set using the score min lt func gt function Even though the user is not required to specify additional alignment options it is often advisable and highly recommended to do so The default parameters are not very strict and as a consequence will also not run very swiftly To see a full list of options please type bismark help on the command line or see the Appendix at the end of this User Guide Directional BS Seq libraries former option directional Bisulfite treatment of DNA and subsequent PCR amplification can give rise to four bisulfite converted strands for a given locus Depending on the adapters used BS Seq libraries can be constructed in two different ways a If a library is directional only reads which are bisulfite converted versions of the original top strand OT or the original bottom strand OB will be sequenced Even though the strands complementary to OT CTOT and OB CTOB are generated in the BS PCR step they will not be sequenced as they carry the wrong kind of adapter at their 5 end By default Bismark performs only 2 read alignments to the OT and
33. the D and R options to adjust the effort expended to find valid alignments For reference this used to be the old now deprecated description of M Bowtie 2 searches for at most lt int gt 1 distinct valid alignments for each read The search terminates when it can t find more distinct valid alignments or when it finds lt int gt 1 distinct alignments whichever happens first Only the best alignment is reported Information from the other alignments is used to estimate mapping quality and to set SAM optional fields such as AS i and XS i Increasing M makes Bowtie 2 slower but increases the likelihood that it will pick the correct alignment for a read that aligns many places for reads that have more than lt int gt 1 distinct valid alignments Bowtie 2 does not guarantee that the alignment reported is the best possible in terms of alignment score M is always used and its default value is set to 10 IlI Bismark Methylation Extractor A brief description of the Bismark methylation extractor and a full list of options can also be viewed by typing methylation extractor help USAGE methylation extractor options lt filenames gt ARGUMENTS lt filenames gt OPTIONS s single end p paired end vanilla no_overlap ignore lt int gt comprehensive merge non CpG A space separated list of result files in Bismark format from which methylation information is extracted for every cytosine in t
34. uch post processing scripts are not part of the Bismark package Bisulfite conversion related non bisulfite mismatches are mismatch positions which have a C in the BS read but a T in the genome such mismatches may occur due to the way bisulfite read alignments are performed Reads containing this kind of mismatches are not automatically removed from the alignment output in order not to introduce a bias for methylated reads please send me an email if you have further questions about this It should be noted that even though no methylation calls are performed for these positions reads containing bisulfite conversion related non bisulfite mismatches might lead to false alignments if particularly lax alignment parameters were specified Methylation call The methylation call string contains a dot for every position in the BS read not involving a cytosine or contains one of the following letters for the three different cytosine methylation contexts UPPER CASE METHYLATED lower case unmethylated unmethylated C in CpG context methylated C in CpG context unmethylated C in CHG context methylated C in CHG context unmethylated C in CHH context methylated C in CHH context rca x xX NN IlI Bismark methylation extractor Bismark comes with a supplementary methylation extractor script which operates on Bismark result files and extracts the methylation call for every single C analysed The position of every single C will be written
35. umber of chromosomes is vast this is likely to be in the range of several thousand files operating systems can only handle lists up to a certain length Some newly assembled genomes may contain 20000 50000 contig of scaffold files which do exceed this list length limit The path to the folder containing the genome to be bisulfite converted this may be an absolute or relative path Bismark Genome Preparation expects one or more FastaA files in the folder valid file extensions fa or fasta If the path is not provided as an argument you will be prompted for it later II Bismark A brief description of Bismark and a full list of options can also be viewed by typing bismark help USAGE bismark lt singles gt ARGUMENTS lt genome_folder gt 1 lt mates1 gt 2 lt mates2 gt lt singles gt OPTIONS Input q fastq f fasta s skip lt int gt u qupto lt int gt options lt genome folder gt 1 lt matesl gt 2 lt mates2 gt The full path to the folder containing the unmodified reference genome as well as the subfolders created by the bismark genome preparation script Bisulfite Genome CT conversion and Bisulfite Genome GA_conversion Bismark expects one or more FastA files in this folder file extension fa or fasta The path to the genome folder can be relative or absolute Comma separated list of files containing the 1 mates filename usually includes _1
Download Pdf Manuals
Related Search