|
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 April 30, 2019
|
|
jpayne@69
|
7
|
|
jpayne@69
|
8 Description: Performs set operations on VCF files:
|
|
jpayne@69
|
9 Union, intersection, and subtraction.
|
|
jpayne@69
|
10
|
|
jpayne@69
|
11 Usage: comparevcf.sh in=<file,file,...> out=<file>
|
|
jpayne@69
|
12
|
|
jpayne@69
|
13 I/O parameters:
|
|
jpayne@69
|
14 in=<file> Input; must be at least 2 files.
|
|
jpayne@69
|
15 out=<file> Output file.
|
|
jpayne@69
|
16 ref=<file> Reference file; optional. Usually not needed.
|
|
jpayne@69
|
17 shist=<file> (scorehist) Output for variant score histogram.
|
|
jpayne@69
|
18 overwrite=f (ow) Set to false to force the program to abort rather than
|
|
jpayne@69
|
19 bgzip=f Use bgzip for gzip compression.
|
|
jpayne@69
|
20
|
|
jpayne@69
|
21 Processing Mode (choose one only):
|
|
jpayne@69
|
22 subtract=t Subtract all other files from the first file.
|
|
jpayne@69
|
23 union=f Make a union of all files.
|
|
jpayne@69
|
24 intersection=f Make an intersection of all files.
|
|
jpayne@69
|
25
|
|
jpayne@69
|
26 Processing Parameters:
|
|
jpayne@69
|
27 addsamples=t Include all samples in the output lines. (TODO)
|
|
jpayne@69
|
28 splitalleles=f Split multi-allelic lines into multiple lines.
|
|
jpayne@69
|
29 splitsubs=f Split multi-base substitutions into SNPs.
|
|
jpayne@69
|
30 canonize=t Trim variations down to a canonical representation.
|
|
jpayne@69
|
31
|
|
jpayne@69
|
32 Java Parameters:
|
|
jpayne@69
|
33 -Xmx This will set Java's memory usage, overriding autodetection.
|
|
jpayne@69
|
34 -Xmx20g will specify 20 gigs of RAM, and -Xmx200m will
|
|
jpayne@69
|
35 specify 200 megs. The max is typically 85% of physical memory.
|
|
jpayne@69
|
36 -eoom This flag will cause the process to exit if an out-of-memory
|
|
jpayne@69
|
37 exception occurs. Requires Java 8u92+.
|
|
jpayne@69
|
38 -da Disable assertions.
|
|
jpayne@69
|
39
|
|
jpayne@69
|
40 Please contact Brian Bushnell at bbushnell@lbl.gov if you encounter any problems.
|
|
jpayne@69
|
41 "
|
|
jpayne@69
|
42 }
|
|
jpayne@69
|
43
|
|
jpayne@69
|
44 #This block allows symlinked shellscripts to correctly set classpath.
|
|
jpayne@69
|
45 pushd . > /dev/null
|
|
jpayne@69
|
46 DIR="${BASH_SOURCE[0]}"
|
|
jpayne@69
|
47 while [ -h "$DIR" ]; do
|
|
jpayne@69
|
48 cd "$(dirname "$DIR")"
|
|
jpayne@69
|
49 DIR="$(readlink "$(basename "$DIR")")"
|
|
jpayne@69
|
50 done
|
|
jpayne@69
|
51 cd "$(dirname "$DIR")"
|
|
jpayne@69
|
52 DIR="$(pwd)/"
|
|
jpayne@69
|
53 popd > /dev/null
|
|
jpayne@69
|
54
|
|
jpayne@69
|
55 #DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )/"
|
|
jpayne@69
|
56 CP="$DIR""current/"
|
|
jpayne@69
|
57
|
|
jpayne@69
|
58 z="-Xmx4g"
|
|
jpayne@69
|
59 z2="-Xms4g"
|
|
jpayne@69
|
60 set=0
|
|
jpayne@69
|
61
|
|
jpayne@69
|
62 if [ -z "$1" ] || [[ $1 == -h ]] || [[ $1 == --help ]]; then
|
|
jpayne@69
|
63 usage
|
|
jpayne@69
|
64 exit
|
|
jpayne@69
|
65 fi
|
|
jpayne@69
|
66
|
|
jpayne@69
|
67 calcXmx () {
|
|
jpayne@69
|
68 source "$DIR""/calcmem.sh"
|
|
jpayne@69
|
69 setEnvironment
|
|
jpayne@69
|
70 parseXmx "$@"
|
|
jpayne@69
|
71 if [[ $set == 1 ]]; then
|
|
jpayne@69
|
72 return
|
|
jpayne@69
|
73 fi
|
|
jpayne@69
|
74 freeRam 4000m 84
|
|
jpayne@69
|
75 z="-Xmx${RAM}m"
|
|
jpayne@69
|
76 z2="-Xms${RAM}m"
|
|
jpayne@69
|
77 }
|
|
jpayne@69
|
78 calcXmx "$@"
|
|
jpayne@69
|
79
|
|
jpayne@69
|
80 comparevcf() {
|
|
jpayne@69
|
81 local CMD="java $EA $EOOM $z $z2 -cp $CP var2.CompareVCF $@"
|
|
jpayne@69
|
82 echo $CMD >&2
|
|
jpayne@69
|
83 eval $CMD
|
|
jpayne@69
|
84 }
|
|
jpayne@69
|
85
|
|
jpayne@69
|
86 comparevcf "$@"
|
