Mercurial > repos > rliterman > csp2

annotate CSP2/CSP2_env/env-d9b9114564458d9d-741b3de822f2aaca6c6caa4325c4afce/opt/bbmap-39.01-1/decontaminate.sh @ 69:33d812a61356

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload commit 2e9511a184a1ca667c7be0c6321a36dc4e3d116d
author jpayne
date Tue, 18 Mar 2025 17:55:14 -0400
parents
children
rev   line source
jpayne@69 1 #!/bin/bash
jpayne@69 2
jpayne@69 3 usage(){
jpayne@69 4 echo "
jpayne@69 5 Written by Brian Bushnell.
jpayne@69 6 Last modified June 28, 2016
jpayne@69 7
jpayne@69 8 Description: Decontaminates multiplexed assemblies via normalization and mapping.
jpayne@69 9
jpayne@69 10 Usage: decontaminate.sh reads=<file,file> ref=<file,file> out=<directory>
jpayne@69 11 or
jpayne@69 12 decontaminate.sh readnamefile=<file> refnamefile=<file> out=<directory>
jpayne@69 13
jpayne@69 14 Input Parameters:
jpayne@69 15 reads=<file,file> Input reads, one file per library.
jpayne@69 16 ref=<file,file> Input assemblies, one file per library.
jpayne@69 17 readnamefile=<file> List of input reads, one line per library.
jpayne@69 18 refnamefile=<file> List of input assemblies, one line per library.
jpayne@69 19
jpayne@69 20 interleaved=auto True forces paired/interleaved input; false forces single-ended mapping.
jpayne@69 21 If not specified, interleaved status will be autodetected from read names.
jpayne@69 22 unpigz=t Spawn a pigz (parallel gzip) process for faster decompression. Requires pigz to be installed.
jpayne@69 23 touppercase=t (tuc) Convert lowercase letters in reads to upper case (otherwise they will not match the reference).
jpayne@69 24
jpayne@69 25 Output Parameters:
jpayne@69 26 pigz=f Spawn a pigz (parallel gzip) process for faster compression. Requires pigz to be installed.
jpayne@69 27 tmpdir=. Write temp files here. By default is uses the system's $TMPDIR or current directory.
jpayne@69 28 outdir=. Write ouput files here.
jpayne@69 29
jpayne@69 30 Mapping Parameters:
jpayne@69 31 kfilter=55 Set to a positive number N to require minimum N contiguous matches for a mapped read.
jpayne@69 32 ambig=random Determines how coverage will be calculated for ambiguously-mapped reads.
jpayne@69 33 first: Add coverage only at first genomic mapping location.
jpayne@69 34 random: Add coverage at a random best-scoring location.
jpayne@69 35 all: Add coverage at all best-scoring locations.
jpayne@69 36 toss: Discard ambiguously-mapped reads without adding coverage.
jpayne@69 37
jpayne@69 38 Filtering Parameters:
jpayne@69 39 minc=3.5 Min average coverage to retain scaffold.
jpayne@69 40 minp=20 Min percent coverage to retain scaffold.
jpayne@69 41 minr=18 Min mapped reads to retain scaffold.
jpayne@69 42 minl=500 Min length to retain scaffold.
jpayne@69 43 ratio=1.2 Contigs will not be removed by minc unless the coverage changed by at least this factor. 0 disables this filter.
jpayne@69 44 mapraw=t Set true to map the unnormalized reads. Required to filter by 'ratio'.
jpayne@69 45 basesundermin=-1 If positive, removes contigs with at least this many bases in low-coverage windows.
jpayne@69 46 window=500 Sliding window size
jpayne@69 47 windowcov=5 Average coverage below this will be classified as low.
jpayne@69 48
jpayne@69 49 Tadpole Parameters:
jpayne@69 50 ecct=f Error-correct with Tadpole before normalization.
jpayne@69 51 kt=42 Kmer length for Tadpole.
jpayne@69 52 aggressive=f Do aggressive error correction.
jpayne@69 53 conservative=f Do conservative error correction.
jpayne@69 54 tadpoleprefilter=1 (tadpre) Ignore kmers under this depth to save memory.
jpayne@69 55
jpayne@69 56 Normalization Parameters:
jpayne@69 57 mindepth=2 Min depth of reads to keep.
jpayne@69 58 target=20 Target normalization depth.
jpayne@69 59 hashes=4 Number of hashes in Bloom filter.
jpayne@69 60 passes=1 Normalization passes.
jpayne@69 61 minprob=0.5 Min probability of correctness to add a kmer.
jpayne@69 62 dp=0.75 (depthpercentile) Percentile to use for depth proxy (0.5 means median).
jpayne@69 63 prefilter=t Prefilter, for large datasets.
jpayne@69 64 filterbits=32 (fbits) Bits per cell in primary filter.
jpayne@69 65 prefilterbits=2 (pbits) Bits per cell in prefilter.
jpayne@69 66 k=31 Kmer length for normalization. Longer is more precise but less sensitive.
jpayne@69 67
jpayne@69 68 Other parameters:
jpayne@69 69 opfn=0 (onlyprocessfirstn) Set to a positive number to only process that many datasets. This is for internal testing of specificity.
jpayne@69 70
jpayne@69 71 Java Parameters:
jpayne@69 72 -Xmx This will set Java's memory usage, overriding autodetection.
jpayne@69 73 -Xmx20g will specify 20 gigs of RAM, and -Xmx800m will specify 800 megs.
jpayne@69 74 The max is typically 85% of physical memory.
jpayne@69 75 -eoom This flag will cause the process to exit if an
jpayne@69 76 out-of-memory exception occurs. Requires Java 8u92+.
jpayne@69 77 -da Disable assertions.
jpayne@69 78
jpayne@69 79 Please contact Brian Bushnell at bbushnell@lbl.gov if you encounter any problems.
jpayne@69 80 "
jpayne@69 81 }
jpayne@69 82
jpayne@69 83 #This block allows symlinked shellscripts to correctly set classpath.
jpayne@69 84 pushd . > /dev/null
jpayne@69 85 DIR="${BASH_SOURCE[0]}"
jpayne@69 86 while [ -h "$DIR" ]; do
jpayne@69 87 cd "$(dirname "$DIR")"
jpayne@69 88 DIR="$(readlink "$(basename "$DIR")")"
jpayne@69 89 done
jpayne@69 90 cd "$(dirname "$DIR")"
jpayne@69 91 DIR="$(pwd)/"
jpayne@69 92 popd > /dev/null
jpayne@69 93
jpayne@69 94 #DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )/"
jpayne@69 95 CP="$DIR""current/"
jpayne@69 96 JNI="-Djava.library.path=""$DIR""jni/"
jpayne@69 97 JNI=""
jpayne@69 98
jpayne@69 99 z="-Xmx1g"
jpayne@69 100 z2="-Xms1g"
jpayne@69 101 set=0
jpayne@69 102
jpayne@69 103 if [ -z "$1" ] || [[ $1 == -h ]] || [[ $1 == --help ]]; then
jpayne@69 104 usage
jpayne@69 105 exit
jpayne@69 106 fi
jpayne@69 107
jpayne@69 108 calcXmx () {
jpayne@69 109 source "$DIR""/calcmem.sh"
jpayne@69 110 setEnvironment
jpayne@69 111 parseXmx "$@"
jpayne@69 112 if [[ $set == 1 ]]; then
jpayne@69 113 return
jpayne@69 114 fi
jpayne@69 115 freeRam 15000m 84
jpayne@69 116 z="-Xmx${RAM}m"
jpayne@69 117 z2="-Xms${RAM}m"
jpayne@69 118 }
jpayne@69 119 calcXmx "$@"
jpayne@69 120
jpayne@69 121
jpayne@69 122 decontaminate() {
jpayne@69 123 local CMD="java $JNI $EA $EOOM $z $z2 -cp $CP jgi.DecontaminateByNormalization $@"
jpayne@69 124 echo $CMD >&2
jpayne@69 125 eval $CMD
jpayne@69 126 }
jpayne@69 127
jpayne@69 128 decontaminate "$@"