6a48e07c2fbefbf0a0e5e74662dd768ef93fca88 hiram Wed Apr 29 11:27:30 2020 -0700 adding ncbiRmsk option to use NCBI rm.out.gz file instead of cluster run and record versionInfo.txt of RepeatMasker operation refs #23734 diff --git src/hg/utils/automation/doAssemblyHub.pl src/hg/utils/automation/doAssemblyHub.pl index 0534908..41b0111 100755 --- src/hg/utils/automation/doAssemblyHub.pl +++ src/hg/utils/automation/doAssemblyHub.pl @@ -16,30 +16,31 @@ use strict; use FindBin qw($Bin); use lib "$Bin"; use HgAutomate; use HgRemoteScript; use HgStepManager; # Option variable names, both common and peculiar to this script: use vars @HgAutomate::commonOptionVars; use vars @HgStepManager::optionVars; use vars qw/ $opt_buildDir $opt_sourceDir $opt_species $opt_rmskSpecies + $opt_ncbiRmsk $opt_augustusSpecies $opt_xenoRefSeq $opt_ucscNames $opt_asmHubName /; # Specify the steps supported with -continue / -stop: my $stepper = new HgStepManager( [ { name => 'download', func => \&doDownload }, { name => 'sequence', func => \&doSequence }, { name => 'assemblyGap', func => \&doAssemblyGap }, { name => 'gatewayPage', func => \&doGatewayPage }, { name => 'cytoBand', func => \&doCytoBand }, { name => 'gc5Base', func => \&doGc5Base }, { name => 'repeatMasker', func => \&doRepeatMasker }, @@ -54,30 +55,32 @@ { name => 'ncbiGene', func => \&doNcbiGene }, { name => 'ncbiRefSeq', func => \&doNcbiRefSeq }, { name => 'xenoRefGene', func => \&doXenoRefGene }, { name => 'augustus', func => \&doAugustus }, { name => 'trackDb', func => \&doTrackDb }, { name => 'cleanup', func => \&doCleanup }, ] ); # Option defaults: my $dbHost = 'hgwdev'; my $sourceDir = "/hive/data/outside/ncbi/genomes"; my $species = ""; # usually found in asmId_assembly_report.txt my $ftpDir = ""; # will be determined from given asmId my $rmskSpecies = ""; +my $ncbiRmsk = 0; # when =1 call doRepeatMasker.pl + # with -ncbiRmsk=path.out.gz and -liftSpec=... my $augustusSpecies = "human"; my $xenoRefSeq = "/hive/data/genomes/asmHubs/VGP/xenoRefSeq"; my $ucscNames = 0; # default 'FALSE' (== 0) my $asmHubName = "n/a"; # directory name in: /gbdb/hubs/asmHubName my $workhorse = "hgwdev"; # default workhorse when none chosen my $fileServer = "hgwdev"; # default when none chosen my $bigClusterHub = "ku"; # default when none chosen my $smallClusterHub = "ku"; # default when none chosen my $base = $0; $base =~ s/^(.*\/)?//; # key is original accession name from the remove.dups.list, value is 1 my %dupAccessionList; @@ -96,30 +99,32 @@ print STDERR $stepper->getOptionHelp(); print STDERR <<_EOF_ -buildDir dir Construct assembly hub in dir instead of default $HgAutomate::clusterData/asmHubs/refseqBuild/GC[AF]/123/456/789/asmId/ -sourceDir dir Find assembly in dir instead of default: $sourceDir/GC[AF]/123/456/789/asmId -ucscNames Translate NCBI/INSDC/RefSeq names to UCSC names default is to use the given NCBI/INSDC/RefSeq names -asmHubName <name> directory name in: /gbdb/hubs/asmHubName -species <name> use this species designation if there is no asmId_assembly_report.txt with an 'Organism name:' entry to obtain species -rmskSpecies <name> to override default 'species' name for repeat masker the default is found in the asmId_asssembly_report.txt e.g. -rmskSpecies=viruses + -ncbiRmsk use NCBI rm.out.gz file instead of local cluster run + for repeat masking -augustusSpecies <human|chicken|zebrafish> default 'human' -xenoRefSeq </path/to/xenoRefSeqMrna> - location of xenoRefMrna.fa.gz expanded directory of mrnas/ and xenoRefMrna.sizes, default $xenoRefSeq _EOF_ ; print STDERR &HgAutomate::getCommonOptionHelp('dbHost' => $dbHost, 'workhorse' => $workhorse, 'fileServer' => $fileServer, 'bigClusterHub' => $bigClusterHub, 'smallClusterHub' => $smallClusterHub); print STDERR " Automates build of assembly hub. Steps: download: sets up sym link working hierarchy from already mirrored files from NCBI in: @@ -174,30 +179,31 @@ } # Globals: # Command line args: asmId my ( $asmId); # Other: my ($buildDir, $secondsStart, $secondsEnd, $assemblySource); sub checkOptions { # Make sure command line options are valid/supported. my $ok = GetOptions(@HgStepManager::optionSpec, 'buildDir=s', 'sourceDir=s', 'rmskSpecies=s', + 'ncbiRmsk', 'augustusSpecies=s', 'xenoRefSeq=s', 'asmHubName=s', 'ucscNames', @HgAutomate::commonOptionSpec, ); &usage(1) if (!$ok); &usage(0, 1) if ($opt_help); &HgAutomate::processCommonOptions(); my $err = $stepper->processOptions(); usage(1) if ($err); $dbHost = $opt_dbHost if ($opt_dbHost); } ######################################################################### @@ -618,36 +624,38 @@ my $runDir = "$buildDir/download"; &HgAutomate::mustMkdir($runDir); my $whatItDoes = "setup work hierarchy of sym links to source files in\n\t$runDir/"; my $bossScript = newBash HgRemoteScript("$runDir/doDownload.bash", $workhorse, $runDir, $whatItDoes); $bossScript->add(<<_EOF_ export asmId=$asmId if [ ! -s \${asmId}.2bit -o \${asmId}_genomic.fna.gz -nt \$asmId.2bit ]; then rm -f \${asmId}_genomic.fna.gz \\ \${asmId}_genomic.fna.dups.gz \\ \${asmId}_assembly_report.txt \\ \${asmId}_rm.out.gz \\ + \${asmId}_rm.run \\ \${asmId}_assembly_structure \\ \$asmId.2bit ln -s $assemblySource/\${asmId}_genomic.fna.gz . ln -s $assemblySource/\${asmId}_assembly_report.txt . ln -s $assemblySource/\${asmId}_rm.out.gz . + ln -s $assemblySource/\${asmId}_rm.run . if [ -d $assemblySource/\${asmId}_assembly_structure ]; then ln -s $assemblySource/\${asmId}_assembly_structure . fi faToTwoBit \${asmId}_genomic.fna.gz \$asmId.2bit twoBitDup \$asmId.2bit > \$asmId.dups.txt if [ -s "\$asmId.dups.txt" ]; then printf "WARNING duplicate sequences found in \$asmId.2bit\\n" 1>&2 cat \$asmId.dups.txt 1>&2 awk '{print \$1}' \$asmId.dups.txt > \$asmId.remove.dups.list mv \${asmId}_genomic.fna.gz \${asmId}_genomic.fna.dups.gz faSomeRecords -exclude \${asmId}_genomic.fna.dups.gz \\ \$asmId.remove.dups.list stdout | gzip -c > \${asmId}_genomic.fna.gz rm -f \$asmId.2bit faToTwoBit \${asmId}_genomic.fna.gz \$asmId.2bit fi @@ -1031,44 +1039,65 @@ sub doRepeatMasker { my $runDir = "$buildDir/trackData/repeatMasker"; if ( -d "$buildDir/trackData/repeatMasker/run.cluster" ) { if ( ! -s "$buildDir/trackData/repeatMasker/faSize.rmsk.txt" ) { &HgAutomate::verbose(1, "\nERROR: step repeatmasker may be running\n"); exit 255; } } &HgAutomate::mustMkdir($runDir); my $whatItDoes = "construct repeatMasker track data"; my $bossScript = newBash HgRemoteScript("$runDir/doRepeatMasker.bash", $workhorse, $runDir, $whatItDoes); + my $rmskOpts = ""; + if ($ncbiRmsk) { + $rmskOpts = " \\ + -ncbiRmsk=\"$buildDir/download/${asmId}_rm.out.gz\" "; + if ($ucscNames) { + $rmskOpts .= " \\ + -liftSpec=\"$buildDir/sequence/$asmId.ncbiToUcsc.lift\""; + } + } + $bossScript->add(<<_EOF_ export asmId=$asmId if [ $buildDir/\$asmId.2bit -nt faSize.rmsk.txt ]; then export species=`echo $rmskSpecies | sed -e 's/_/ /g;'` -doRepeatMasker.pl -stop=mask -buildDir=`pwd` -unmaskedSeq=$buildDir/\$asmId.2bit \\ +rm -f versionInfo.txt + +doRepeatMasker.pl -stop=mask -buildDir=`pwd` -unmaskedSeq=$buildDir/\$asmId.2bit $rmskOpts \\ -bigClusterHub=$bigClusterHub -workhorse=$workhorse -species="\$species" \$asmId -gzip \$asmId.sorted.fa.out \$asmId.fa.out \$asmId.nestedRepeats.bed +if [ -s "\$asmId.fa.out" ]; then + gzip \$asmId.fa.out +fi +gzip \$asmId.sorted.fa.out \$asmId.nestedRepeats.bed -doRepeatMasker.pl -continue=cleanup -buildDir=`pwd` -unmaskedSeq=$buildDir/\$asmId.2bit \\ +doRepeatMasker.pl -continue=cleanup -buildDir=`pwd` -unmaskedSeq=$buildDir/\$asmId.2bit $rmskOpts \\ -bigClusterHub=$bigClusterHub -workhorse=$workhorse -species="\$species" \$asmId +if [ ! -s versionInfo.txt ]; then + if [ -s ../../download/${asmId}_rm.run ]; then + ln -s ../../download/${asmId}_rm.run versionInfo.txt + fi +fi + \$HOME/kent/src/hg/utils/automation/asmHubRepeatMasker.sh \$asmId `pwd`/\$asmId.sorted.fa.out.gz `pwd` else printf "# repeatMasker step previously completed\\n" 1>&2 exit 0 fi _EOF_ ); $bossScript->execute(); } # repeatMasker ######################################################################### # * step: simpleRepeat [workhorse] sub doSimpleRepeat { my $runDir = "$buildDir/trackData/simpleRepeat"; &HgAutomate::mustMkdir($runDir); @@ -1531,37 +1560,53 @@ bigBedInfo \$asmId.ncbiGene.bb | egrep "^itemCount:|^basesCovered:" \\ | sed -e 's/,//g' > \$asmId.ncbiGene.stats.txt LC_NUMERIC=en_US /usr/bin/printf "# ncbiGene %s %'d %s %'d\\n" `cat \$asmId.ncbiGene.stats.txt` | xargs echo printf "%d bases of %d (%s%%) in intersection\\n" "\$baseCount" "\$asmSizeNoGaps" "\$perCent" > fb.\$asmId.ncbiGene.txt else printf "# ncbiGene step previously completed\\n" 1>&2 fi _EOF_ ); $bossScript->execute(); } # doNcbiGene ######################################################################### # * step: ncbiRefSeq [workhorse] sub doNcbiRefSeq { - my $runDir = "$buildDir/trackData/ncbiRefSeq"; + # skip this procedure if all the required files are not available my $gffFile = "$assemblySource/${asmId}_genomic.gff.gz"; if ( ! -s "${gffFile}" ) { printf STDERR "# step ncbiRefSeq no gff file found at:\n# %s\n", $gffFile; return; } + my $filesFound = 0; + my @requiredFiles = qw( genomic.gff.gz rna.fna.gz rna.gbff.gz protein.faa.gz ); + my $filesExpected = scalar(@requiredFiles); + foreach my $expectFile (@requiredFiles) { + if ( -s "$assemblySource/${asmId}_${expectFile}" ) { + ++$filesFound; + } else { + printf STDERR "# step ncbiRefSeq missing required file $assemblySource/${asmId}_${expectFile}\n"; + } + } + if ($filesFound < $filesExpected) { + printf STDERR "# step ncbiRefSeq does not have all files required\n"; + return; + } + + my $runDir = "$buildDir/trackData/ncbiRefSeq"; &HgAutomate::mustMkdir($runDir); my $whatItDoes = "run NCBI RefSeq gene procedures"; my $bossScript = newBash HgRemoteScript("$runDir/doNcbiRefSeq.bash", $workhorse, $runDir, $whatItDoes); $bossScript->add(<<_EOF_ export asmId="$asmId" export buildDir="$buildDir" export liftFile="\$buildDir/sequence/\$asmId.ncbiToUcsc.lift" export target2bit="\$buildDir/\$asmId.2bit" if [ $buildDir/\$asmId.2bit -nt \$asmId.ncbiRefSeq.bb ]; then ~/kent/src/hg/utils/automation/doNcbiRefSeq.pl -toGpWarnOnly -buildDir=`pwd` \\ @@ -1736,31 +1781,31 @@ } if (length($species) < 1) { die "no -species specified and can not find Organism name: in $asmReport"; } } $rmskSpecies = $opt_rmskSpecies ? $opt_rmskSpecies : $species; $augustusSpecies = $opt_augustusSpecies ? $opt_augustusSpecies : $augustusSpecies; $xenoRefSeq = $opt_xenoRefSeq ? $opt_xenoRefSeq : $xenoRefSeq; $ucscNames = $opt_ucscNames ? 1 : $ucscNames; # '1' == 'TRUE' $workhorse = $opt_workhorse ? $opt_workhorse : $workhorse; $bigClusterHub = $opt_bigClusterHub ? $opt_bigClusterHub : $bigClusterHub; $smallClusterHub = $opt_smallClusterHub ? $opt_smallClusterHub : $smallClusterHub; $fileServer = $opt_fileServer ? $opt_fileServer : $fileServer; $asmHubName = $opt_asmHubName ? $opt_asmHubName : $asmHubName; - +$ncbiRmsk = $opt_ncbiRmsk ? 1 : 0; die "can not find assembly source directory\n$assemblySource" if ( ! -d $assemblySource); printf STDERR "# buildDir: %s\n", $buildDir; printf STDERR "# sourceDir %s\n", $sourceDir; printf STDERR "# augustusSpecies %s\n", $augustusSpecies; printf STDERR "# xenoRefSeq %s\n", $xenoRefSeq; printf STDERR "# assemblySource: %s\n", $assemblySource; printf STDERR "# asmHubName %s\n", $asmHubName; printf STDERR "# rmskSpecies %s\n", $rmskSpecies; printf STDERR "# augustusSpecies %s\n", $augustusSpecies; # Do everything. $stepper->execute(); # Tell the user anything they should know.