9.3 Parameters on Reads

9.3.1 Error models

See estimateErrors on how the post-mortem damamge models and recal models are estimated and formatted.

Parameter Description Default
--RGInfo rginfo.json File defining PMD and quality recalibration parameters. no PMD & recal.
--pmd pmd_model Specify Post-mortem damage parameters. Will overwrite pmd-model defined with --RGInfo if given. no PMD
--recal recal_model Specify Quality score recalibration parameters. Will overwrite pmd-model defined with --RGInfo if given. no recal

9.3.2 Range Filters

  • Range format: [min, max] where min and max are included or ]min, max[ or (min, max) where min and max re excluded. If no brackets are given, the values are included, min,max = [min, max]
  • Ommiting min means 0: ,max] = [0, max], omitting max means inf: ]3, = ]3,inf[
  • If min and max are integers, absolute values are filtered: readLength, fragmentLength, number of softclipped bases, …
  • If min and max are probabilites, relative values are filtered (when it makes sense): number of softclipped bases/readLength, number of mapping mismatches/fragmentLength, …
  • In some cases, the range can be omitted, meaning filter all

Example

  • --filterSoftClips Filters all reads having softclipped bases
  • --filterSoftClips ",3" Filters all reads having more than 3 softclipped bases
  • --filterSoftClips ",0.1" Filters all reads where the softclipped bases represent more than 10% of all bases
Parameter Description Default
--filterMappingLength range Restrict the length of the mapped reads to the range. [0,500]
--filterReadLength range Restrict the Read length to the range. keep all
--filterMQ range Restrict the mapping quality of the reads to the range. keep all
--filterFragmentLength range Restrict the Fragment length to the range. keep all
--filterPMDS range Restrict the PMDS score to the range. keep all
--filterSoftClips range Filter out softclipped reads where the softclipped lengtht is outside the range. keep all
--filterIndels range Filter out reads where the length of insertions and deletions is outside the range. keep all
--filterMappingMismatches range Filter out reads where the mapping mismatches are outside of range. keep all

9.3.3 Binary Filters/Keepers

Parameter Description Default
--keepDuplicates Keep duplicate reads. filter out
--keepImproperPairs Keep Improper pairs. filter out
--keepUnmappedReads Keep Unmapped reads. filter out
--keepFailedQC Keep reads that Failed QC. filter out
--keepSecondaryReads Keep Secondary reads. filter out
--keepSupplementaryReads Keep Supplementary reads. filter out
--keepOnlyFwd Keep only forward reads. Both are kept
--keepOnlyRev Keep only reverse reads. Both are kept
--keepOnlyFirst Keep only the first Mate. Both are kept
--keepOnlySecond Keep only the second Mate. Both are kept
--keepReadsWithoutRG Keep reads without Readgroups. filter out
--filterReadsLongerThanFragment Filter out reads longer than fragment size. keep all
--filterWithAlternativeHits Filter out reads with alternative hits. keep all

9.3.4 Filter Lists

Parameter Description Default
--blacklist blacklist.txt Filter out specific reads provided as a list. keep all
--readGroup list Only use reads belonging to list of Readgroups. keep all
--downsampleReads prob Downsample reads to given probability. keep all

9.3.5 Trimming

Parameter Description Default
--trim5 value trim the first value bases on the 5’ end. don’t trim
--trim3 value trim the first value bases on the 3’ end. don’t trim

9.3.6 Duplicates

Usually, reads are already flagged as duplcates from third-party programs. However, ATlAS also has a duplicates identification algorithm.

--resetDuplicates Remove duplicate flags already in read description.
--markDuplicates Identify duplicates and mark one of the reads.