Tools and Paths

[wmaguire-gladstone]

MonkeyPipeline/bananas_agw.pm

Lines 738 to 791 in e010cd6

	sub setupBin($) {
	my ($cfg) = @_;
	(scalar(@_) == 1) or confess "Wrong number of args to setupBin. The only argument must be the '$cfg' pointer to the hash where we store everything.";
	# Anything in $cfg->{bin} gets checked to make sure it both EXISTS and also CAN BE EXECUTED by the current user.
	# Do NOT put non-executables in $cfg->{bin}, or the automatic checking will fail!
	defined($cfg->{monkeyPoo}) or confess "[ERROR in code]: 'monkeyPoo' directory must have already been set up in the 'checkConfig' function BEFORE the binary directories can be set up.";
	my $binDir = $GLOBAL_BIN_DIR;
	my $MONKEY_SUPPORT_SUBDIR = catfile($cfg->{monkeyPoo}, "support");
	$cfg->{bin}->{basedir} = $binDir;
	$cfg->{bin}->{convertToGenomeBrowserExe} = catfile($binDir,"convert_SAM_or_BAM_for_Genome_Browser.pl");
	$cfg->{bin}->{qsub} = catfile($binDir,"qsub"); # queue submission utility
	$cfg->{bin}->{qhold} = catfile($binDir,"qhold"); # queue 'hold' utility
	$cfg->{bin}->{qrls} = catfile($binDir,"qrls"); # queue 'release' utility (opposite of qhold)
	$cfg->{bin}->{bam2bed} = catfile($binDir,"bam2bed");
	$cfg->{bin}->{bigWig} = catfile($binDir,"bedGraphToBigWig");
	$cfg->{bin}->{bedmap} = catfile($binDir,"bedmap");
	$cfg->{bin}->{bedops} = catfile($binDir,"bedops");
	$cfg->{bin}->{bigBed} = catfile($binDir,"bedToBigBed");
	$cfg->{bin}->{bedtools} = catfile($binDir,"bedtools");
	$cfg->{bin}->{bowtie2} = catfile($binDir,"bowtie2");
	$cfg->{bin}->{bwa} = catfile($binDir,"bwa");
	$cfg->{bin}->{STAR} = $cfg->{bin}->{star}= catfile($binDir,"STAR"); # <== allow both upper AND lower case for the STAR aligner, which is officially capitalized.
	# $cfg->{bin}->{hisat2_0_1} = catfile($binDir,"hisat_2_0_1"); # for requesting a specific version
	# $cfg->{bin}->{hisat} = catfile($binDir,"hisat_current");
	# $cfg->{bin}->{tophat2_1_1} = catfile($binDir,"tophat_2_1_1"); # for requesting a specific version
	$cfg->{bin}->{tophat} = catfile($binDir,"tophat");
	$cfg->{bin}->{cuffdiff} = catfile($binDir,"cuffdiff");
	$cfg->{bin}->{fastqMcf} = catfile($binDir,"fastq-mcf");
	$cfg->{bin}->{fastqc} = catfile($binDir,"fastqc");
	$cfg->{bin}->{gem} = catfile($binDir,"gem.jar");
	$cfg->{bin}->{bcp} = catfile($binDir,"BCP_HM");
	$cfg->{bin}->{match} = catfile($binDir,"match");
	$cfg->{bin}->{match2moa} = catfile($binDir,"match2moa");
	$cfg->{bin}->{moaOverlaps} = catfile($binDir,"removeMoaOverlaps");
	$cfg->{bin}->{samtools} = catfile($binDir,"samtools");
	$cfg->{bin}->{sortBed} = catfile($binDir,"sort-bed");
	$cfg->{bin}->{subreadFeatureCounts} = catfile($binDir,"featureCounts");
	$cfg->{bin}->{featureCountsToPlainMatrix}= catfile($MONKEY_SUPPORT_SUBDIR, "featureCounts_to_plain_matrix.sh");
	$cfg->{bin}->{alt_analyze_py} = catfile($binDir,"AltAnalyze.py");
	$cfg->{bin}->{peakMotifs} = "/data/applications/rsat/rsat/perl-scripts/peak-motifs"; # must be hardcoded to work... dubious. Also requires Perl library "MIME::Lite"
	$cfg->{bin}->{R_EXE} = "/usr/bin/R";
	$cfg->{bin}->{RSCRIPT_EXE} = "/usr/bin/Rscript"; # Some sub-scripts (like 51c.cluster.R) need this, you but should not need to run it manually.
	$cfg->{bin}->{python2_7_exe} = "/bin/python2.7"; # location of a python executable with version >= 2.7
	$cfg->{bin}->{atacPython} = "/home/sthomas/envs/atac/bin/python";
	$cfg->{bin}->{nucleoatac} = "/home/sthomas/envs/atac/bin/nucleoatac";
	$cfg->{bin}->{step1Motif} = catfile($cfg->{monkeyPoo},"s1_getMotifRegex.pl"); # /home/sthomas/tools/utils/s1_getMotifRegex.pl
	$cfg->{bin}->{step2Motif} = catfile($cfg->{monkeyPoo},"s2_findLocations.pl"); # /home/sthomas/tools/utils/s2_findLocations.pl
	}

There's a list of tools/commands - we'll need to make sure to have these all installed.

Path on line 780, 786, 787

[AlexanderPico]

You may have done already, right?

[wmaguire-gladstone]

On Rigel there's a directory /data/bin (that's a symbolic link to /data/applications/2015_06/bin/) that has all these tools, except in /data/bin they're symbolic links to /data/applications/APPLICATIONS_BASE/toolname.

Seems like it was work towards creating an area where you could update the apps without breaking everyones path. We should discuss this, my proposal is:
/wynton/group/gladstone/third_party/toolname/current/bin that is sym linked to /toolname/versionnumber/bin
Then add to the PATH variable as: /wynton/group/gladstone/third_party//current/bin
Then also create specified ones like /third_party//binf_current, or dev, binf_dev, and then just a quick PATH update to change toolsets.

[wmaguire-gladstone]

Started an rsync of the tools from Rigel, on Monday Nov 4. Once in place, will go through each tool and make necessary changes and updates.

[wmaguire-gladstone]

Updated line of peak-motifs.

The lines referring to atacPython and nucleoatac are referencing to an area that doesn't exist in Rigel, I've commented them out for now.

The rest of the tools should be called correctly if BINFBINROOT is set correctly - see issue #29