|
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 January 28, 2020
|
|
jpayne@69
|
7
|
|
jpayne@69
|
8 Description: Merges files of SSU sequences to keep one per taxID.
|
|
jpayne@69
|
9 By default, a consensus is generated per TaxID, then the sequence
|
|
jpayne@69
|
10 best matching that consensus is used:
|
|
jpayne@69
|
11 First, all sequences per TaxID are aligned to a reference consensus.
|
|
jpayne@69
|
12 Second, the best-matching sequence is used as a seed, and all other
|
|
jpayne@69
|
13 sequences for that TaxID are aligned to the seed to generate a new consensus.
|
|
jpayne@69
|
14 Third, in 'consensus' mode, that consensus is simply output.
|
|
jpayne@69
|
15 In 'best' mode (default), all sequences are aligned again to the new consensus,
|
|
jpayne@69
|
16 and the best-matching is output.
|
|
jpayne@69
|
17
|
|
jpayne@69
|
18 Usage: mergeribo.sh in=<file,file> out=<file> 16S
|
|
jpayne@69
|
19
|
|
jpayne@69
|
20 Standard parameters:
|
|
jpayne@69
|
21 in=<file,file> Comma-delimited list of files.
|
|
jpayne@69
|
22 out=<file> Output file.
|
|
jpayne@69
|
23 out2=<file> Read 2 output if reads are in two files.
|
|
jpayne@69
|
24 overwrite=f (ow) Set to false to force the program to abort rather than
|
|
jpayne@69
|
25 overwrite an existing file.
|
|
jpayne@69
|
26 showspeed=t (ss) Set to 'f' to suppress display of processing speed.
|
|
jpayne@69
|
27 ziplevel=2 (zl) Set to 1 (lowest) through 9 (max) to change compression
|
|
jpayne@69
|
28 level; lower compression is faster.
|
|
jpayne@69
|
29
|
|
jpayne@69
|
30 Processing parameters:
|
|
jpayne@69
|
31 alt=<file> Lower priority data. Only used if there is no SSU associated
|
|
jpayne@69
|
32 with the TaxID from the primary input.
|
|
jpayne@69
|
33 best=t Output the best representative per taxID.
|
|
jpayne@69
|
34 consensus=f Output a consensus per taxID instead of the best input
|
|
jpayne@69
|
35 sequence. Mutually exclusive with best.
|
|
jpayne@69
|
36 fast=f Output the best sequence based on alignment to global consensus
|
|
jpayne@69
|
37 (the seed) rather than individual consensus.
|
|
jpayne@69
|
38 minid=0.62 Ignore sequences with identity lower than this to the global
|
|
jpayne@69
|
39 consensus.
|
|
jpayne@69
|
40 maxns=-1 Ignore sequences with more than this many Ns, if non-negative.
|
|
jpayne@69
|
41 minlen=1 Ignore sequences shorter than this.
|
|
jpayne@69
|
42 maxlen=4000 Ignore sequences longer than this.
|
|
jpayne@69
|
43 16S=t Align to 16S consensus to pick the seed. Mutually exclusive.
|
|
jpayne@69
|
44 18S=f Align to 18S consensus to pick the seed. Mutually exclusive.
|
|
jpayne@69
|
45
|
|
jpayne@69
|
46 Java Parameters:
|
|
jpayne@69
|
47 -Xmx This will set Java's memory usage, overriding autodetection.
|
|
jpayne@69
|
48 -Xmx20g will specify 20 gigs of RAM, and -Xmx200m will
|
|
jpayne@69
|
49 specify 200 megs. The max is typically 85% of physical memory.
|
|
jpayne@69
|
50 -eoom This flag will cause the process to exit if an out-of-memory
|
|
jpayne@69
|
51 exception occurs. Requires Java 8u92+.
|
|
jpayne@69
|
52 -da Disable assertions.
|
|
jpayne@69
|
53
|
|
jpayne@69
|
54 Please contact Brian Bushnell at bbushnell@lbl.gov if you encounter any problems.
|
|
jpayne@69
|
55 "
|
|
jpayne@69
|
56 }
|
|
jpayne@69
|
57
|
|
jpayne@69
|
58 #This block allows symlinked shellscripts to correctly set classpath.
|
|
jpayne@69
|
59 pushd . > /dev/null
|
|
jpayne@69
|
60 DIR="${BASH_SOURCE[0]}"
|
|
jpayne@69
|
61 while [ -h "$DIR" ]; do
|
|
jpayne@69
|
62 cd "$(dirname "$DIR")"
|
|
jpayne@69
|
63 DIR="$(readlink "$(basename "$DIR")")"
|
|
jpayne@69
|
64 done
|
|
jpayne@69
|
65 cd "$(dirname "$DIR")"
|
|
jpayne@69
|
66 DIR="$(pwd)/"
|
|
jpayne@69
|
67 popd > /dev/null
|
|
jpayne@69
|
68
|
|
jpayne@69
|
69 #DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )/"
|
|
jpayne@69
|
70 CP="$DIR""current/"
|
|
jpayne@69
|
71
|
|
jpayne@69
|
72 z="-Xmx4g"
|
|
jpayne@69
|
73 z2="-Xms4g"
|
|
jpayne@69
|
74 set=0
|
|
jpayne@69
|
75
|
|
jpayne@69
|
76 if [ -z "$1" ] || [[ $1 == -h ]] || [[ $1 == --help ]]; then
|
|
jpayne@69
|
77 usage
|
|
jpayne@69
|
78 exit
|
|
jpayne@69
|
79 fi
|
|
jpayne@69
|
80
|
|
jpayne@69
|
81 calcXmx () {
|
|
jpayne@69
|
82 source "$DIR""/calcmem.sh"
|
|
jpayne@69
|
83 setEnvironment
|
|
jpayne@69
|
84 parseXmx "$@"
|
|
jpayne@69
|
85 if [[ $set == 1 ]]; then
|
|
jpayne@69
|
86 return
|
|
jpayne@69
|
87 fi
|
|
jpayne@69
|
88 freeRam 4000m 42
|
|
jpayne@69
|
89 z="-Xmx${RAM}m"
|
|
jpayne@69
|
90 z2="-Xms${RAM}m"
|
|
jpayne@69
|
91 }
|
|
jpayne@69
|
92 calcXmx "$@"
|
|
jpayne@69
|
93
|
|
jpayne@69
|
94 mergeribo() {
|
|
jpayne@69
|
95 local CMD="java $EA $EOOM $z -cp $CP prok.MergeRibo $@"
|
|
jpayne@69
|
96 echo $CMD >&2
|
|
jpayne@69
|
97 eval $CMD
|
|
jpayne@69
|
98 }
|
|
jpayne@69
|
99
|
|
jpayne@69
|
100 mergeribo "$@"
|
