|
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 July 31, 2018
|
|
jpayne@69
|
7
|
|
jpayne@69
|
8 Description: Subsamples reads to reach a target unique kmer limit.
|
|
jpayne@69
|
9
|
|
jpayne@69
|
10 Differences between versions:
|
|
jpayne@69
|
11 kmerlimit.sh uses 1 pass and outputs all reads until a limit is hit,
|
|
jpayne@69
|
12 meaning the input reads should be in random order with respect to sequence.
|
|
jpayne@69
|
13 kmerlimit2.sh uses 2 passes and randomly subsamples from the file, so
|
|
jpayne@69
|
14 it works with reads in any order.
|
|
jpayne@69
|
15
|
|
jpayne@69
|
16 Usage: kmerlimit2.sh in=<input file> out=<output file> limit=<number>
|
|
jpayne@69
|
17
|
|
jpayne@69
|
18 Standard parameters:
|
|
jpayne@69
|
19 in=<file> Primary input, or read 1 input.
|
|
jpayne@69
|
20 in2=<file> Read 2 input if reads are in two files.
|
|
jpayne@69
|
21 out=<file> Primary output, or read 1 output.
|
|
jpayne@69
|
22 out2=<file> Read 2 output if reads are in two files.
|
|
jpayne@69
|
23 overwrite=t (ow) Set to false to force the program to abort rather than
|
|
jpayne@69
|
24 overwrite an existing file.
|
|
jpayne@69
|
25 ziplevel=2 (zl) Set to 1 (lowest) through 9 (max) to change compression
|
|
jpayne@69
|
26 level; lower compression is faster.
|
|
jpayne@69
|
27
|
|
jpayne@69
|
28 Processing parameters:
|
|
jpayne@69
|
29 k=31 Kmer length, 1-32.
|
|
jpayne@69
|
30 limit= The number of unique kmers to produce.
|
|
jpayne@69
|
31 mincount=1 Ignore kmers seen fewer than this many times.
|
|
jpayne@69
|
32 minqual=0 Ignore bases with quality below this.
|
|
jpayne@69
|
33 minprob=0.2 Ignore kmers with correctness probability below this.
|
|
jpayne@69
|
34 trials=25 Simulation trials.
|
|
jpayne@69
|
35 seed=-1 Set to a positive number for deterministic output.
|
|
jpayne@69
|
36 maxlen=50m Max length of a temp array used in simulation.
|
|
jpayne@69
|
37
|
|
jpayne@69
|
38 Java Parameters:
|
|
jpayne@69
|
39 -Xmx This will set Java's memory usage, overriding autodetection.
|
|
jpayne@69
|
40 -Xmx20g will specify 20 gigs of RAM, and -Xmx200m will
|
|
jpayne@69
|
41 specify 200 megs. The max is typically 85% of physical memory.
|
|
jpayne@69
|
42 -eoom This flag will cause the process to exit if an out-of-memory
|
|
jpayne@69
|
43 exception occurs. Requires Java 8u92+.
|
|
jpayne@69
|
44 -da Disable assertions.
|
|
jpayne@69
|
45
|
|
jpayne@69
|
46 Please contact Brian Bushnell at bbushnell@lbl.gov if you encounter any problems.
|
|
jpayne@69
|
47 "
|
|
jpayne@69
|
48 }
|
|
jpayne@69
|
49
|
|
jpayne@69
|
50 #This block allows symlinked shellscripts to correctly set classpath.
|
|
jpayne@69
|
51 pushd . > /dev/null
|
|
jpayne@69
|
52 DIR="${BASH_SOURCE[0]}"
|
|
jpayne@69
|
53 while [ -h "$DIR" ]; do
|
|
jpayne@69
|
54 cd "$(dirname "$DIR")"
|
|
jpayne@69
|
55 DIR="$(readlink "$(basename "$DIR")")"
|
|
jpayne@69
|
56 done
|
|
jpayne@69
|
57 cd "$(dirname "$DIR")"
|
|
jpayne@69
|
58 DIR="$(pwd)/"
|
|
jpayne@69
|
59 popd > /dev/null
|
|
jpayne@69
|
60
|
|
jpayne@69
|
61 #DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )/"
|
|
jpayne@69
|
62 CP="$DIR""current/"
|
|
jpayne@69
|
63
|
|
jpayne@69
|
64 z="-Xmx1000m"
|
|
jpayne@69
|
65 z2="-Xms1000m"
|
|
jpayne@69
|
66 set=0
|
|
jpayne@69
|
67
|
|
jpayne@69
|
68 if [ -z "$1" ] || [[ $1 == -h ]] || [[ $1 == --help ]]; then
|
|
jpayne@69
|
69 usage
|
|
jpayne@69
|
70 exit
|
|
jpayne@69
|
71 fi
|
|
jpayne@69
|
72
|
|
jpayne@69
|
73 calcXmx () {
|
|
jpayne@69
|
74 source "$DIR""/calcmem.sh"
|
|
jpayne@69
|
75 setEnvironment
|
|
jpayne@69
|
76 parseXmx "$@"
|
|
jpayne@69
|
77 }
|
|
jpayne@69
|
78 calcXmx "$@"
|
|
jpayne@69
|
79
|
|
jpayne@69
|
80 kmerlimit2() {
|
|
jpayne@69
|
81 local CMD="java $EA $EOOM $z -cp $CP sketch.KmerLimit2 $@"
|
|
jpayne@69
|
82 echo $CMD >&2
|
|
jpayne@69
|
83 eval $CMD
|
|
jpayne@69
|
84 }
|
|
jpayne@69
|
85
|
|
jpayne@69
|
86 kmerlimit2 "$@"
|
