Mosdepth uses HTSLib ( ) via the nim programming language ( ) it expects the input BAM or CRAM file to be sorted by position. Here, we introduce mosdepth and show that it is faster than existing methods and has additional utility. The need for efficient coverage calculation increases with the number and depth of whole genome sequences, and existing methods require roughly an hour or more of computation for a typical human genome with 30Г— coverage. Samtools depth ( Li et al., 2009) outputs per-base coverage BEDTools genomecov ( Quinlan and Hall, 2010 Quinlan, 2014) can output per-region or per-base coverage Sambamba ( Tarasov et al., 2015) also provides per-base and per-window depth calculations. Given the scope of applications for coverage profiles, there are several existing tools that calculate genome-wide coverage. by cn.mops ( Klambauer et al., 2012), quality control ( Pedersen et al., 2017), and determining which genomic regions have too low, or too high ( Li, 2014) coverage for reliable variant calling. Written by Andrew Whitwham from the Sanger Institute.Measuring the depth of sequencing coverage is critical for genomic analyses such as calling copy-number variation (CNV), e.g. Thus no additional sort steps are normally needed. Sequence alignment and the markdup step after sorting by chromosome and Typically the fixmate step would be applied immediately after Samtools markdup positionsort.bam markdup.bam Samtools sort -o positionsort.bam fixmate.bam Samtools fixmate -m namecollate.bam fixmate.bam Samtools collate -o namecollate.bam example.bamДЄdd ms and MC tags for markdup to use later: This first collate command can be omitted if the file is already TheyД¬an optionally be marked as duplicates if they have a primary that is also a Excluded reads are not used for calculating duplicates. By default QC failed reads are also excluded but can be includedДЄs an option. Might be obtained by further sequencing of the library.Д®xcluded reads are those marked as secondary, supplementary or In particular it can be used to determine how much more data However itД¬an provide a useful guide into how many unique read pairs are likely to beДЄvailable. the libraryД¬onsists of unique fragments that are randomly selected (with replacement) SINGLE: reads that are not part of a pair.ДUPLICATE PAIR: reads in a duplicate pair.ДUPLICATE SINGLE: single read duplicates.ДUPLICATE PAIR OPTICAL: optical duplicate paired reads.ДUPLICATE SINGLE OPTICAL: optical duplicate single reads.ДUPLICATE NON PRIMARY: supplementary/secondary duplicate reads.ДUPLICATE NON PRIMARY OPTICAL: supplementary/secondary opticalДUPLICATE PRIMARY TOTAL: number of primary duplicate reads.ДUPLICATE TOTAL: total number of duplicate reads.Д®STIMATED LIBRARY SIZE: estimate of the number of unique fragments inД®stimated library size makes various assumptions e.g. threads INT Number of input/output compression threads to use in addition to mainД®XAMINED: reads examined for duplication. no-PG Do not add a PG line to the output file. When there are a great many duplicates for each original read. Using this option can speed up duplicate marking Tagging where reads may tagged with a better quality read but not Marking reads as duplicates further checks to make sure all opticalДuplicates are found are not carried out. no-multi-dup Stop checking duplicates of duplicates for correctness. include-fails Include quality checked failed reads. The two methods identify mostly the same reads as duplicates, mode

Mode s measures positions based on sequence start. Mode t measures positions based on template start/end TYPE Duplicate decision method for paired reads. t Mark duplicates with the name of the original in a do tag.

c Clear previous duplicate settings and tags. Optical duplicateДetection will not work on non standard names. Names produced by the Illumina sequencing machines. When set, duplicate reads are tagged withДt:Z:SQ for optical duplicates and dt:Z:LB otherwise.Д¬alculation of distance depends on coordinate data embedded in the read Platforms or about 2500 for NovaSeq ones. Suggested settings of 100 for HiSeq style d distance The optical duplicate distance. tmp -S Mark supplementary reads of duplicates as duplicates. OPTIONS В¶ -l INT Expected maximum read length of INT bases. Program relies on the MC and ms tags that fixmate provides. Mark duplicate alignments from a coordinate sorted file that hasД«een run through samtools fixmate with the -m option. Samtools-markdup - mark duplicate alignments in a coordinate

Samtools threads