|
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 March 21, 2019
|
|
jpayne@69
|
7
|
|
jpayne@69
|
8 Description: Calculates observed quality scores from mapped sam/bam files.
|
|
jpayne@69
|
9 Generates matrices for use in recalibrating quality scores. By default,
|
|
jpayne@69
|
10 the matrices are written to /ref/qual/ in the current directory.
|
|
jpayne@69
|
11
|
|
jpayne@69
|
12 If you have multiple sam/bam files demultiplexed from a single sequencing run,
|
|
jpayne@69
|
13 it is recommended to use all of them as input for increased statistical power.
|
|
jpayne@69
|
14 Once the matrices are generated, recalibration can be done on mapped or
|
|
jpayne@69
|
15 unmapped reads; you may get better results by recalibrating the fastq and
|
|
jpayne@69
|
16 remapping the calibrated reads.
|
|
jpayne@69
|
17
|
|
jpayne@69
|
18 Note! Diploid organisms with a high heterozygousity rate will induce
|
|
jpayne@69
|
19 inaccurate recalibration at the high end of the quality scale unless SNP
|
|
jpayne@69
|
20 locations are masked or variations are called. For example, recalibrating
|
|
jpayne@69
|
21 human reads mapped to an unmasked human reference would generate an
|
|
jpayne@69
|
22 expected maximal Q-score of roughly 30 due to the human 1/1000 SNP rate.
|
|
jpayne@69
|
23 Variations can be ignored by using the callvars flag or providing
|
|
jpayne@69
|
24 a file of variations.
|
|
jpayne@69
|
25
|
|
jpayne@69
|
26 Usage:
|
|
jpayne@69
|
27
|
|
jpayne@69
|
28 Step 1. Generate matrices (from mapped sam or bam files):
|
|
jpayne@69
|
29 calctruequality.sh in=<file,file,...file> path=<directory>
|
|
jpayne@69
|
30
|
|
jpayne@69
|
31 Step 2. Recalibrate reads (any kind of files):
|
|
jpayne@69
|
32 bbduk.sh in=<file> out=<file> recalibrate
|
|
jpayne@69
|
33
|
|
jpayne@69
|
34
|
|
jpayne@69
|
35 Parameters (and their defaults)
|
|
jpayne@69
|
36
|
|
jpayne@69
|
37 Input parameters:
|
|
jpayne@69
|
38 in=<file,file> Sam file or comma-delimited list of files. Alignments
|
|
jpayne@69
|
39 must use = and X cigar symbols, or have MD tags, or
|
|
jpayne@69
|
40 ref must be specified.
|
|
jpayne@69
|
41 reads=-1 Stop after processing this many reads (if positive).
|
|
jpayne@69
|
42 samstreamer=t (ss) Load reads multithreaded to increase speed.
|
|
jpayne@69
|
43 unpigz=t Use pigz to decompress.
|
|
jpayne@69
|
44
|
|
jpayne@69
|
45 Output parameters:
|
|
jpayne@69
|
46 overwrite=t (ow) Set to true to allow overwriting of existing files.
|
|
jpayne@69
|
47 path=. Directory to write quality matrices (within /ref subdir).
|
|
jpayne@69
|
48 write=t Write matrices.
|
|
jpayne@69
|
49 showstats=t Print a summary.
|
|
jpayne@69
|
50 pigz=f Use pigz to compress.
|
|
jpayne@69
|
51
|
|
jpayne@69
|
52 Other parameters:
|
|
jpayne@69
|
53 t=auto Number of worker threads.
|
|
jpayne@69
|
54 passes=2 Recalibration passes, 1 or 2. 2 is slower but gives more
|
|
jpayne@69
|
55 accurate quality scores.
|
|
jpayne@69
|
56 recalqmax=42 Adjust max quality scores tracked. The actual highest
|
|
jpayne@69
|
57 quality score allowed is recalqmax-1.
|
|
jpayne@69
|
58 trackall=f Track all available quality metrics and produce all
|
|
jpayne@69
|
59 matrices, including the ones that are not selected for
|
|
jpayne@69
|
60 quality adjustment. Reduces speed, but allows testing the
|
|
jpayne@69
|
61 effects of different recalibration matrices.
|
|
jpayne@69
|
62 indels=t Include indels in quality calculations.
|
|
jpayne@69
|
63
|
|
jpayne@69
|
64 Variation calling:
|
|
jpayne@69
|
65 varfile=<file> Use the variants in this var file, instead of calling
|
|
jpayne@69
|
66 variants. The format can be produced by CallVariants.
|
|
jpayne@69
|
67 vcf=<file> Use the variants in this VCF file, instead of
|
|
jpayne@69
|
68 calling variants.
|
|
jpayne@69
|
69 callvars=f Call SNPs, and do not count them as errors.
|
|
jpayne@69
|
70 ploidy=1 Set the organism's ploidy.
|
|
jpayne@69
|
71 ref= Required for variation-calling.
|
|
jpayne@69
|
72
|
|
jpayne@69
|
73 *** 'Variant-Calling Cutoffs' flags in callvariants.sh are also supported ***
|
|
jpayne@69
|
74
|
|
jpayne@69
|
75 Selecting matrices:
|
|
jpayne@69
|
76 loadq102= For each recalibration matrix, enable or disable that matrix with t/f.
|
|
jpayne@69
|
77 You can specify pass1 or pass2 like this: loadq102_p1=f loadq102_p2=t.
|
|
jpayne@69
|
78 The default is loadqbp_p1=t loadqbp_p2=t loadqb123_p=t.
|
|
jpayne@69
|
79 clearmatrices=f If true, clear all the existing matrix selections. For example:
|
|
jpayne@69
|
80 'clearmatrices loadqbp_p1'
|
|
jpayne@69
|
81 This would ignore defaults and select only qbp for the first pass.
|
|
jpayne@69
|
82
|
|
jpayne@69
|
83 Available matrices:
|
|
jpayne@69
|
84 q102 Quality, leading quality, trailing quality.
|
|
jpayne@69
|
85 qap Quality, average quality, position.
|
|
jpayne@69
|
86 qbp Quality, current base, position.
|
|
jpayne@69
|
87 q10 Quality, leading quality.
|
|
jpayne@69
|
88 q12 Quality, trailing quality.
|
|
jpayne@69
|
89 qb12 Quality, leading base, current base.
|
|
jpayne@69
|
90 qb012 Quality, two leading bases, current base.
|
|
jpayne@69
|
91 qb123 Quality, leading base, current base, trailing base.
|
|
jpayne@69
|
92 qb234 Quality, current base, two trailing bases.
|
|
jpayne@69
|
93 q12b12 Quality, trailing quality, leading base, current base.
|
|
jpayne@69
|
94 qp Quality, position.
|
|
jpayne@69
|
95 q Current quality score only.
|
|
jpayne@69
|
96
|
|
jpayne@69
|
97
|
|
jpayne@69
|
98 Java Parameters:
|
|
jpayne@69
|
99 -Xmx This will set Java's memory usage, overriding autodetection.
|
|
jpayne@69
|
100 -Xmx20g will specify 20 gigs of RAM, and -Xmx200m will specify 200 megs.
|
|
jpayne@69
|
101 The max is typically 85% of physical memory.
|
|
jpayne@69
|
102 -eoom This flag will cause the process to exit if an
|
|
jpayne@69
|
103 out-of-memory exception occurs. Requires Java 8u92+.
|
|
jpayne@69
|
104 -da Disable assertions.
|
|
jpayne@69
|
105
|
|
jpayne@69
|
106 Please contact Brian Bushnell at bbushnell@lbl.gov if you encounter any problems.
|
|
jpayne@69
|
107 "
|
|
jpayne@69
|
108 }
|
|
jpayne@69
|
109
|
|
jpayne@69
|
110 #This block allows symlinked shellscripts to correctly set classpath.
|
|
jpayne@69
|
111 pushd . > /dev/null
|
|
jpayne@69
|
112 DIR="${BASH_SOURCE[0]}"
|
|
jpayne@69
|
113 while [ -h "$DIR" ]; do
|
|
jpayne@69
|
114 cd "$(dirname "$DIR")"
|
|
jpayne@69
|
115 DIR="$(readlink "$(basename "$DIR")")"
|
|
jpayne@69
|
116 done
|
|
jpayne@69
|
117 cd "$(dirname "$DIR")"
|
|
jpayne@69
|
118 DIR="$(pwd)/"
|
|
jpayne@69
|
119 popd > /dev/null
|
|
jpayne@69
|
120
|
|
jpayne@69
|
121 #DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )/"
|
|
jpayne@69
|
122 CP="$DIR""current/"
|
|
jpayne@69
|
123
|
|
jpayne@69
|
124 z="-Xmx2g"
|
|
jpayne@69
|
125 z2="-Xms2g"
|
|
jpayne@69
|
126 set=0
|
|
jpayne@69
|
127
|
|
jpayne@69
|
128 if [ -z "$1" ] || [[ $1 == -h ]] || [[ $1 == --help ]]; then
|
|
jpayne@69
|
129 usage
|
|
jpayne@69
|
130 exit
|
|
jpayne@69
|
131 fi
|
|
jpayne@69
|
132
|
|
jpayne@69
|
133 calcXmx () {
|
|
jpayne@69
|
134 source "$DIR""/calcmem.sh"
|
|
jpayne@69
|
135 setEnvironment
|
|
jpayne@69
|
136 parseXmx "$@"
|
|
jpayne@69
|
137 if [[ $set == 1 ]]; then
|
|
jpayne@69
|
138 return
|
|
jpayne@69
|
139 fi
|
|
jpayne@69
|
140 freeRam 3200m 84
|
|
jpayne@69
|
141 z="-Xmx${RAM}m"
|
|
jpayne@69
|
142 z2="-Xms${RAM}m"
|
|
jpayne@69
|
143 }
|
|
jpayne@69
|
144 calcXmx "$@"
|
|
jpayne@69
|
145
|
|
jpayne@69
|
146 calctruequality() {
|
|
jpayne@69
|
147 local CMD="java $EA $EOOM $z $z2 -cp $CP jgi.CalcTrueQuality $@"
|
|
jpayne@69
|
148 echo $CMD >&2
|
|
jpayne@69
|
149 eval $CMD
|
|
jpayne@69
|
150 }
|
|
jpayne@69
|
151
|
|
jpayne@69
|
152 calctruequality "$@"
|
