# for emacs: -*- mode: sh; -*- # DATE: 11-Apr-2012 # ORGANISM: Geospiza fortis # TAXID: 48883 # ASSEMBLY LONG NAME: GeoFor_1.0 # ASSEMBLY SHORT NAME: GeoFor_1.0 # ASSEMBLY SUBMITTER: Beijing Genomics Institute # ASSEMBLY TYPE: Haploid # NUMBER OF ASSEMBLY-UNITS: 1 # ASSEMBLY ACCESSION: GCA_000277835.1 # FTP-RELEASE DATE: 25-Jul-2012 # http://www.ncbi.nlm.nih.gov/genome/13302 # http://www.ncbi.nlm.nih.gov/genome/assembly/402638/ # http://www.ncbi.nlm.nih.gov/bioproject/156703 # http://www.ncbi.nlm.nih.gov/Traces/wgs/?val=AKZB01 # Genome Coverage : 115x # http://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?id=48883 # rsync://ftp.ncbi.nlm.nih.gov/genbank/genomes/Eukaryotes/vertebrates_other/Geospiza_fortis/GeoFor_1.0/ ########################################################################## # Download sequence (DONE - 2012-07-26 - Hiram) mkdir /hive/data/genomes/geoFor1 cd /hive/data/genomes/geoFor1 mkdir genbank cd genbank rsync -a -P \ rsync://ftp.ncbi.nlm.nih.gov/genbank/genomes/Eukaryotes/vertebrates_other/Geospiza_fortis/GeoFor_1.0/ ./ # verify the size of the sequence here: faSize Primary_Assembly/unplaced_scaffolds/FASTA/unplaced.scaf.fa.gz # 1065292181 bases (24006152 N's 1041286029 real 1041286029 upper 0 lower) # in 27239 sequences in 1 files # Total size: mean 39109.1 sd 545732.4 min 131 (gi|397531944|gb|JH767125.1|) # max 30495534 (gi|397559183|gb|JH739887.1|) median 598 # %0.00 masked total, %0.00 masked real # strip the names down to something reasonable, they can all be: # >JH7[0-9] zcat Primary_Assembly/unplaced_scaffolds/FASTA/unplaced.scaf.fa.gz \ | sed -e "s/^>.*JH7/>JH7/; s/.1. Geospiza.*//" | gzip -c > ucsc.fa.gz zcat Primary_Assembly/unplaced_scaffolds/AGP/unplaced.scaf.agp.gz \ | sed -e "s/^\(JH7[0-9]*\).1/\1/" | gzip -c > ucsc.agp.gz checkAgpAndFa ucsc.agp.gz ucsc.fa.gz 2>&1 | tail -2 # Valid Fasta file entry # All AGP and FASTA entries agree - both files are valid mkdir /hive/data/genomes/geoFor1/photograph cd /hive/data/genomes/geoFor1/photograph wget -O female_galapagos_medium_ground_finch.jpg \ http://upload.wikimedia.org/wikipedia/commons/7/7b/Female_Gal%C3%A1pagos_medium_ground_finch.jpg convert -geometry "350x350" female_galapagos_medium_ground_finch.jpg \ Geospiza_fortis.jpg # check this .jpg file into the source tree kent/src/hg/htdocs/images/ git commit -m "from Wikimedia Commons - Charlesjsharp" Geospiza_fortis.jpg # and copy to /usr/local/apache/htdocs/images cp -p Geospiza_fortis.jpg /usr/local/apache/htdocs/images ########################################################################## # Initial makeGenomeDb.pl (DONE - 2012-07-26 - Hiram) # obtain a template for this from the source tree: # kent/src/hg/utils/automation/configFiles/ # and check it back into the source tree when completed here: cd /hive/data/genomes/geoFor1 cat << '_EOF_' > geoFor1.config.ra # Config parameters for makeGenomeDb.pl: db geoFor1 clade vertebrate genomeCladePriority 60 scientificName Geospiza fortis commonName Medium Ground Finch assemblyDate Apr. 2012 assemblyLabel Beijing Genomics Institute assemblyShortLabel GeoFor_1.0 orderKey 4373 mitoAcc none fastaFiles /hive/data/genomes/geoFor1/genbank/ucsc.fa.gz agpFiles /hive/data/genomes/geoFor1/genbank/ucsc.agp.gz # qualFiles none dbDbSpeciesDir birds photoCreditURL http://commons.wikimedia.org/wiki/User:Charlesjsharp photoCreditName Photo courtesy of Charles J Sharp, Wikimedia Commons ncbiGenomeId 13302 ncbiAssemblyId 402638 ncbiAssemblyName GeoFor_1.0 ncbiBioProject 156703 genBankAccessionID GCA_000277835.1 taxId 48883 '_EOF_' # << happy emacs time makeGenomeDb.pl -workhorse=hgwdev -fileServer=hgwdev -dbHost=hgwdev \ -stop=agp geoFor1.config.ra > agp.log 2>&1 # real 1m4.001s # verify OK: tail -1 agp.log # *** All done! (through the 'agp' step) # finish it off time makeGenomeDb.pl -continue=db -workhorse=hgwdev -fileServer=hgwdev \ -dbHost=hgwdev geoFor1.config.ra > db.log 2>&1 # real 9m24.133s # add the trackDb entries to the source tree, and the 2bit link: ln -s `pwd`/geoFor1.unmasked.2bit /gbdb/geoFor1/geoFor1.2bit # browser should function now ########################################################################## # running repeat masker (DONE - 2012-07-26 - Hiram) mkdir /hive/data/genomes/geoFor1/bed/repeatMasker cd /hive/data/genomes/geoFor1/bed/repeatMasker time doRepeatMasker.pl -buildDir=`pwd` -noSplit \ -bigClusterHub=swarm -dbHost=hgwdev -workhorse=hgwdev \ -smallClusterHub=encodek geoFor1 > do.log 2>&1 & # real 242m59.832s cat faSize.rmsk.txt # 1065292181 bases (24006152 N's 1041286029 real 981275330 upper # 60010699 lower) in 27239 sequences in 1 files # Total size: mean 39109.1 sd 545732.4 min 131 (JH767125) # max 30495534 (JH739887) median 598 # %5.63 masked total, %5.76 masked real egrep -i "versi|relea" do.log # April 26 2011 (open-3-3-0) version of RepeatMasker # CC RELEASE 20110920; # RepeatMasker version development-$Id: RepeatMasker,v 1.26 2011/09/26 16:19:44 angie Exp $ featureBits -countGaps geoFor1 rmsk # 60063362 bases of 1065292181 (5.638%) in intersection # why is it different than the faSize above ? # because rmsk masks out some N's as well as bases, the count above # separates out the N's from the bases, it doesn't show lower case N's ########################################################################## # running simple repeat (DONE - 2012-07-26 - Hiram) mkdir /hive/data/genomes/geoFor1/bed/simpleRepeat cd /hive/data/genomes/geoFor1/bed/simpleRepeat time doSimpleRepeat.pl -buildDir=`pwd` -bigClusterHub=swarm \ -dbHost=hgwdev -workhorse=hgwdev -smallClusterHub=encodek \ geoFor1 > do.log 2>&1 & # real 27m0.573s cat fb.simpleRepeat # 23621674 bases of 1041286029 (2.269%) in intersection ######################################################################### # Verify all gaps are marked, add any N's not in gap as type 'other' # (DONE - 2012-07-26 - Hiram) mkdir /hive/data/genomes/geoFor1/bed/gap cd /hive/data/genomes/geoFor1/bed/gap time nice -n +19 findMotif -motif=gattaca -verbose=4 \ -strand=+ ../../geoFor1.unmasked.2bit > findMotif.txt 2>&1 # real 0m9.384s grep "^#GAP " findMotif.txt | sed -e "s/^#GAP //" > allGaps.bed time featureBits geoFor1 -not gap -bed=notGap.bed # 1041286029 bases of 1041286029 (100.000%) in intersection # real 0m7.332s # can see now if allGaps.bed actually is all the gaps: hgsql -N -e "select size from gap;" geoFor1 | ave stdin | grep total # total 24006152.000000 ave -col=5 allGaps.bed | grep total # total 24006152.000000 # the rest is unnecessary, they are all marked time featureBits geoFor1 allGaps.bed notGap.bed -bed=new.gaps.bed # 0 bases of 1041286029 (0.000%) in intersection # real 10m13.458s # no new gaps, nothing to do here # check if any non-bridged gaps here: hgsql -N -e "select bridge from gap;" geoFor1 | sort | uniq -c # 68589 yes ########################################################################## ## WINDOWMASKER (DONE - 2012-07-26 - Hiram) mkdir /hive/data/genomes/geoFor1/bed/windowMasker cd /hive/data/genomes/geoFor1/bed/windowMasker time nice -n +19 doWindowMasker.pl -buildDir=`pwd` -workhorse=hgwdev \ -dbHost=hgwdev geoFor1 > do.log 2>&1 & # real 56m40.276s # Masking statistics cat faSize.geoFor1.wmsk.txt # 1065292181 bases (24006152 N's 1041286029 real 840644896 upper # 200641133 lower) in 27239 sequences in 1 files # Total size: mean 39109.1 sd 545732.4 min 131 (JH767125) # max 30495534 (JH739887) median 598 # %18.83 masked total, %19.27 masked real cat faSize.geoFor1.wmsk.sdust.txt # 1065292181 bases (24006152 N's 1041286029 real 833616540 upper # 207669489 lower) in 27239 sequences in 1 files # Total size: mean 39109.1 sd 545732.4 min 131 (JH767125) # max 30495534 (JH739887) median 598 # %19.49 masked total, %19.94 masked real cat faSize.geoFor1.cleanWMSdust.txt # 1065292181 bases (24006152 N's 1041286029 real 833616540 upper # 207669489 lower) in 27239 sequences in 1 files # Total size: mean 39109.1 sd 545732.4 min 131 (JH767125) # max 30495534 (JH739887) median 598 # %19.49 masked total, %19.94 masked real cat fb.geoFor1.windowmaskerSdust.clean.txt # 207669489 bases of 1065292181 (19.494%) in intersection # how much does this window masker and repeat masker overlap: featureBits -countGaps geoFor1 rmsk windowmaskerSdust # 32967187 bases of 1065292181 (3.095%) in intersection ########################################################################## # add simpleRepeats to WindowMasker result (DONE - 2012-07-26 - Hiram) # add to rmsk after it is done: cd /hive/data/genomes/geoFor1 twoBitMask -add bed/windowMasker/geoFor1.cleanWMSdust.2bit \ bed/simpleRepeat/trfMask.bed geoFor1.2bit # you can safely ignore the warning about fields >= 13 twoBitToFa geoFor1.2bit stdout | faSize stdin > faSize.geoFor1.2bit.txt cat faSize.geoFor1.2bit.txt # 1065292181 bases (24006152 N's 1041286029 real 833083628 upper # 208202401 lower) in 27239 sequences in 1 files # Total size: mean 39109.1 sd 545732.4 min 131 (JH767125) # max 30495534 (JH739887) median 598 # %19.54 masked total, %19.99 masked real rm /gbdb/geoFor1/geoFor1.2bit ln -s `pwd`/geoFor1.2bit /gbdb/geoFor1/geoFor1.2bit ########################################################################## # cpgIslands - (DONE - 2012-07-26 - Hiram) mkdir /hive/data/genomes/geoFor1/bed/cpgIslands cd /hive/data/genomes/geoFor1/bed/cpgIslands time doCpgIslands.pl geoFor1 > do.log 2>&1 # real 17m6.876s cat fb.geoFor1.cpgIslandExt.txt # 5087816 bases of 1041286029 (0.489%) in intersection ######################################################################### # genscan - (DONE - 2012-07-26 - Hiram) mkdir /hive/data/genomes/geoFor1/bed/genscan cd /hive/data/genomes/geoFor1/bed/genscan time doGenscan.pl geoFor1 > do.log 2>&1 # real 31m32.788s # Completed: 27239 of 27239 jobs # CPU time in finished jobs: 28s 0.47m 0.01h 0.00d 0.000 y # IO & Wait Time: 96830s 1613.83m 26.90h 1.12d 0.003 y # Average job time: 4s 0.06m 0.00h 0.00d # Longest finished job: 16s 0.27m 0.00h 0.00d # Submission to last job: 154s 2.57m 0.04h 0.00d # Estimated complete: 0s 0.00m 0.00h 0.00d cat fb.geoFor1.genscan.txt # 18118133 bases of 1041286029 (1.740%) in intersection cat fb.geoFor1.genscanSubopt.txt # 22582067 bases of 1041286029 (2.169%) in intersection ######################################################################### # MAKE 11.OOC FILE FOR BLAT/GENBANK (DONE - 2012-07-26 - Hiram) # Use -repMatch=400, based on size -- for human we use 1024 # use the "real" number from the faSize measurement, # hg19 is 2897316137, calculate the ratio factor for 1024: calc \( 1041286029 / 2897316137 \) \* 1024 # ( 1041286029 / 2897316137 ) * 1024 = 368.022281 # round up to 400 (melUnd1 was also 400) cd /hive/data/genomes/geoFor1 time blat geoFor1.2bit /dev/null /dev/null -tileSize=11 \ -makeOoc=jkStuff/geoFor1.11.ooc -repMatch=400 # Wrote 14450 overused 11-mers to jkStuff/geoFor1.11.ooc # real 0m32.689s # melUnd1 was: Wrote 14287 overused 11-mers to jkStuff/melUnd1.11.ooc # there are no non-bridged gaps, no lift file needed for genbank hgsql -N -e "select bridge from gap;" geoFor1 | sort | uniq -c # 68589 yes # cd /hive/data/genomes/geoFor1/jkStuff # gapToLift geoFor1 geoFor1.nonBridged.lift -bedFile=geoFor1.nonBridged.bed # largest non-bridged contig: # awk '{print $3-$2,$0}' geoFor1.nonBridged.bed | sort -nr | head # 123773608 chrX 95534 123869142 chrX.01 ######################################################################### # AUTO UPDATE GENBANK (DONE - 2012-07-26 - Hiram) # examine the file: /cluster/data/genbank/data/organism.lst # for your species to see what counts it has for: # organism mrnaCnt estCnt refSeqCnt # Geospiza fortis 1 0 0 # to decide which "native" mrna or ests you want to specify in genbank.conf ssh hgwdev cd $HOME/kent/src/hg/makeDb/genbank git pull # edit etc/genbank.conf to add geoFor1 just after ce2 # geoFor1 (Medium Ground Finch - Darwin finch) geoFor1.serverGenome = /hive/data/genomes/geoFor1/geoFor1.2bit geoFor1.clusterGenome = /hive/data/genomes/geoFor1/geoFor1.2bit geoFor1.ooc = /hive/data/genomes/geoFor1/jkStuff/geoFor1.11.ooc geoFor1.lift = no geoFor1.refseq.mrna.native.pslCDnaFilter = ${lowCover.refseq.mrna.native.pslCDnaFilter} geoFor1.refseq.mrna.xeno.pslCDnaFilter = ${lowCover.refseq.mrna.xeno.pslCDnaFilter} geoFor1.genbank.mrna.native.pslCDnaFilter = ${lowCover.genbank.mrna.native.pslCDnaFilter} geoFor1.genbank.mrna.xeno.pslCDnaFilter = ${lowCover.genbank.mrna.xeno.pslCDnaFilter} geoFor1.genbank.est.native.pslCDnaFilter = ${lowCover.genbank.est.native.pslCDnaFilter} geoFor1.refseq.mrna.native.load = no geoFor1.refseq.mrna.xeno.load = yes geoFor1.genbank.mrna.xeno.load = no geoFor1.genbank.est.native.load = no geoFor1.downloadDir = geoFor1 geoFor1.perChromTables = no # end of section added to etc/genbank.conf git commit -m "adding geoFor1 medium ground finch" etc/genbank.conf git push make etc-update git pull # Edit src/lib/gbGenome.c to add new species. git commit -m "adding definition for geoForNames" src/lib/gbGenome.c git push make install-server ssh hgwdev # used to do this on "genbank" machine screen -S geoFor1 # long running job managed in screen cd /cluster/data/genbank time nice -n +19 ./bin/gbAlignStep -initial geoFor1 & # var/build/logs/2012.07.26-22:50:11.geoFor1.initalign.log # real 175m22.294s # load database when finished ssh hgwdev cd /cluster/data/genbank time nice -n +19 ./bin/gbDbLoadStep -drop -initialLoad geoFor1 & # real 9m57.487s # var/dbload/hgwdev/logs/2012.07.27-14:05:35.dbload.log # check the end of that dbload.log to see if it was successful # hgwdev 2012.07.27-14:15:33 dbload: finish # enable daily alignment and update of hgwdev (DONE - 2012-05-09 - Hiram) cd ~/kent/src/hg/makeDb/genbank git pull # add geoFor1 to: vi etc/align.dbs etc/hgwdev.dbs git commit -m "Added geoFor1." etc/align.dbs etc/hgwdev.dbs git push make etc-update ######################################################################### # set default position to FOXP2 gene displays (DONE - 2012-08-02 - Hiram) hgsql -e \ 'update dbDb set defaultPos="JH739914:135428-435957" where name="geoFor1";' \ hgcentraltest ############################################################################ # downloads and pushQ entry (DONE - 2012-07-27 - Hiram) # after adding geoFor1 to the all.joiner file and verifying that # joinerCheck is clean, can construct the downloads: cd /hive/data/genomes/geoFor1 time makeDownloads.pl -workhorse=hgwdev geoFor1 # real 11m40.799s mkdir /hive/data/genomes/geoFor1/pushQ cd /hive/data/genomes/geoFor1/pushQ # Mark says don't let the transMap track get there time makePushQSql.pl geoFor1 2> stderr.txt | grep -v transMap > geoFor1.sql # real 3m52.778s # check the stderr.txt for bad stuff, these kinds of warnings are OK: # WARNING: hgwdev does not have /gbdb/geoFor1/wib/gc5Base.wib # WARNING: hgwdev does not have /gbdb/geoFor1/wib/quality.wib # WARNING: hgwdev does not have /gbdb/geoFor1/bbi/quality.bw # WARNING: geoFor1 does not have seq # WARNING: geoFor1 does not have extFile # WARNING: geoFor1 does not have estOrientInfo scp -p geoFor1.sql hgwbeta:/tmp ssh hgwbeta "hgsql qapushq < /tmp/geoFor1.sql" ########################################################################## # BLATSERVERS ENTRY (DONE - 2012-08-03 - Hiram) # After getting a blat server assigned by the Blat Server Gods, ssh hgwdev hgsql -e 'INSERT INTO blatServers (db, host, port, isTrans, canPcr) \ VALUES ("geoFor1", "blat4d", "17806", "1", "0"); \ INSERT INTO blatServers (db, host, port, isTrans, canPcr) \ VALUES ("geoFor1", "blat4d", "17807", "0", "1");' \ hgcentraltest # test it with some sequence ############################################################################ # lastz melUnd1 Budgerigar (DONE - 2012-02-29 - Hiram) # original alignment cd /hive/data/genomes/melUnd1/bed/lastzGeoFor1.2012-07-28 cat fb.melUnd1.chainGeoFor1Link.txt # 832321696 bases of 1086614815 (76.598%) in intersection # and this swap: mkdir /hive/data/genomes/geoFor1/bed/blastz.melUnd1.swap cd /hive/data/genomes/geoFor1/bed/blastz.melUnd1.swap time nice -n +19 doBlastzChainNet.pl -verbose=2 \ /hive/data/genomes/melUnd1/bed/lastzGeoFor1.2012-07-28/DEF \ -swap -syntenicNet \ -workhorse=hgwdev -smallClusterHub=encodek -bigClusterHub=swarm \ > swap.log 2>&1 & # real 56m32.909s cat fb.geoFor1.chainMelUnd1Link.txt # 839798431 bases of 1041286029 (80.650%) in intersection # set sym link to indicate this is the lastz for this genome: cd /hive/data/genomes/geoFor1/bed ln -s blastz.melUnd1.swap lastz.melUnd1 # lets see what recipBest looks like cd /hive/data/genomes/geoFor1/bed/blastz.melUnd1.swap time doRecipBest.pl geoFor1 melUnd1 -buildDir=`pwd` -workhorse=hgwdev \ > best.log 2>&1 & # real 31m15.597s cd /hive/data/genomes/geoFor1/bed/blastz.melUnd1.swap/axtChain hgLoadChain -test -noBin -tIndex geoFor1 \ chainTestRBest geoFor1.melUnd1.rbest.chain.gz # which can be measured with featureBits: mv link.tab rbest.link.bed hgLoadBed geoFor1 tLinkMelUnd1 rbest.link.bed # Read 16927102 elements of size 5 from rbest.link.bed time featureBits geoFor1 tLinkMelUnd1 # 799972446 bases of 1041286029 (76.825%) in intersection # real 1m42.662s hgLoadChain -test -noBin -tIndex geoFor1 \ chainMelGal1RBest geoFor1.melUnd1.syn.chain.gz mv link.tab syntenic.link.bed hgLoadBed geoFor1 tLinkMelUnd1 syntenic.link.bed # Read 16427533 elements of size 5 from syntenic.link.bed time featureBits geoFor1 tLinkMelUnd1 # 787091148 bases of 1041286029 (75.588%) in intersection # real 1m35.489s # cleanup hgsql -e "drop table tLinkMelUnd1;" geoFor1 rm chain.tab bed.tab ############################################################################## # lastz hg19 Human (DONE - 2012-02-29 - Hiram) # original alignment to human cd /hive/data/genomes/hg19/bed/lastzGeoFor1.2012-07-29 cat fb.hg19.chainGeoFor1Link.txt # 101503916 bases of 2897316137 (3.503%) in intersection # and for this swap mkdir /hive/data/genomes/geoFor1/bed/blastz.hg19.swap cd /hive/data/genomes/geoFor1/bed/blastz.hg19.swap time nice -n +19 doBlastzChainNet.pl -verbose=2 \ /hive/data/genomes/hg19/bed/lastzGeoFor1.2012-07-29/DEF \ -workhorse=hgwdev -smallClusterHub=encodek -bigClusterHub=swarm \ -swap -chainMinScore=5000 -chainLinearGap=loose > swap.log 2>&1 & # real 9m10.240s cat fb.geoFor1.chainHg19Link.txt # 88547518 bases of 1041286029 (8.504%) in intersection # set sym link to indicate this is the lastz for this genome: cd /hive/data/genomes/geoFor1/bed ln -s blastz.hg19.swap lastz.hg19 # lets see what synNets look like: cd /hive/data/genomes/geoFor1/bed/blastz.hg19.swap time nice -n +19 doBlastzChainNet.pl -verbose=2 -syntenicNet \ /hive/data/genomes/hg19/bed/lastzGeoFor1.2012-07-29/DEF \ -workhorse=hgwdev -smallClusterHub=encodek -bigClusterHub=swarm \ -swap -continue=syntenicNet -chainMinScore=5000 -chainLinearGap=loose \ > synNet.log 2>&1 & # real 2m28.817s # lets see what recipBest looks like cd /hive/data/genomes/geoFor1/bed/blastz.hg19.swap time doRecipBest.pl geoFor1 hg19 -buildDir=`pwd` -workhorse=hgwdev \ > best.log 2>&1 & # real 8m27.775s # to measure the chain coverage, this test load leaves the file link.tab; hgLoadChain -test -noBin -tIndex geoFor1 \ chainHg19RBest geoFor1.hg19.rbest.chain.gz # which can be measured with featureBits: mv link.tab rbest.link.bed hgLoadBed geoFor1 tLinkHg19 rbest.link.bed # Read 2323598 elements of size 5 from link.bed time featureBits geoFor1 tLinkHg19 # 83552974 bases of 1041286029 (8.024%) in intersection # real 0m20.469s hgLoadChain -test -noBin -tIndex geoFor1 \ chainHg19RBest geoFor1.hg19.syn.chain.gz mv link.tab syntenic.link.bed hgLoadBed geoFor1 tLinkHg19 syntenic.link.bed # Read 2012071 elements of size 5 from syntenic.link.bed time featureBits geoFor1 tLinkHg19 # 74395958 bases of 1041286029 (7.145%) in intersection # real 0m20.854s # cleanup hgsql -e "drop table tLinkHg19;" geoFor1 rm chain.tab bed.tab ######################################################################### # lastz mm10 Mouse (DONE - 2012-02-29 - Hiram) # original alignment to mouse: cd /hive/data/genomes/mm10/bed/lastzGeoFor1.2012-07-29 cat fb.mm10.chainGeoFor1Link.txt # 93984241 bases of 2652783500 (3.543%) in intersection # and for this swap mkdir /hive/data/genomes/geoFor1/bed/blastz.mm10.swap cd /hive/data/genomes/geoFor1/bed/blastz.mm10.swap time nice -n +19 doBlastzChainNet.pl -verbose=2 \ /hive/data/genomes/mm10/bed/lastzGeoFor1.2012-07-29/DEF \ -workhorse=hgwdev -smallClusterHub=encodek -bigClusterHub=swarm \ -swap -chainMinScore=5000 -chainLinearGap=loose > swap.log 2>&1 & # real 10m0.875s cat fb.geoFor1.chainMm10Link.txt # 80273915 bases of 1041286029 (7.709%) in intersection # set sym link to indicate this is the lastz for this genome: cd /hive/data/genomes/geoFor1/bed ln -s blastz.mm10.swap lastz.mm10 # lets see what synNets look like: cd /hive/data/genomes/geoFor1/bed/blastz.mm10.swap time nice -n +19 doBlastzChainNet.pl -verbose=2 -syntenicNet \ /hive/data/genomes/mm10/bed/lastzGeoFor1.2012-07-29/DEF \ -workhorse=hgwdev -smallClusterHub=encodek -bigClusterHub=swarm \ -swap -continue=syntenicNet -chainMinScore=5000 -chainLinearGap=loose \ > synNet.log 2>&1 & # real 2m7.016s # lets see what recipBest looks like cd /hive/data/genomes/geoFor1/bed/blastz.mm10.swap time doRecipBest.pl geoFor1 mm10 -buildDir=`pwd` -workhorse=hgwdev \ > best.log 2>&1 & # real 11m32.814s cd /hive/data/genomes/geoFor1/bed/blastz.mm10.swap/axtChain hgLoadChain -test -noBin -tIndex geoFor1 \ chainMm10RBest geoFor1.mm10.rbest.chain.gz # which can be measured with featureBits: mv link.tab rbest.link.bed hgLoadBed geoFor1 tLinkMm10 rbest.link.bed # Read 1956367 elements of size 5 from rbest.link.bed time featureBits geoFor1 tLinkMm10 # 72866028 bases of 1041286029 (6.998%) in intersection # real 0m18.893s hgLoadChain -test -noBin -tIndex geoFor1 \ chainMm10RBest geoFor1.mm10.syn.chain.gz mv link.tab syntenic.link.bed hgLoadBed geoFor1 tLinkMm10 syntenic.link.bed # Read 1572202 elements of size 5 from syntenic.link.bed time featureBits geoFor1 tLinkMm10 # 61422245 bases of 1041286029 (5.899%) in intersection # real 0m15.096s # cleanup hgsql -e "drop table tLinkMm10;" geoFor1 rm chain.tab bed.tab ######################################################################### # lastz melGal1 Turkey (DONE - 2012-02-29 - Hiram) # original alignment to turkey: cd /hive/data/genomes/melGal1/bed/lastzGeoFor1.2012-07-28 cat fb.melGal1.chainGeoFor1Link.txt # 679715264 bases of 935922386 (72.625%) in intersection # and for this swap mkdir /hive/data/genomes/geoFor1/bed/blastz.melGal1.swap cd /hive/data/genomes/geoFor1/bed/blastz.melGal1.swap time nice -n +19 doBlastzChainNet.pl -verbose=2 \ /hive/data/genomes/melGal1/bed/lastzGeoFor1.2012-07-28/DEF \ -swap -syntenicNet \ -workhorse=hgwdev -smallClusterHub=encodek -bigClusterHub=swarm \ > swap.log 2>&1 & # real 43m59.718s cat fb.geoFor1.chainHg19Link.txt # 698483400 bases of 1041286029 (67.079%) in intersection # set sym link to indicate this is the lastz for this genome: cd /hive/data/genomes/geoFor1/bed ln -s blastz.melGal1.swap lastz.melGal1 # lets see what recipBest looks like cd /hive/data/genomes/geoFor1/bed/blastz.melGal1.swap time doRecipBest.pl geoFor1 melGal1 -buildDir=`pwd` -workhorse=hgwdev \ > best.log 2>&1 & # real 27m31.386s cd /hive/data/genomes/geoFor1/bed/blastz.melGal1.swap/axtChain hgLoadChain -test -noBin -tIndex geoFor1 \ chainMelGal1RBest geoFor1.melGal1.rbest.chain.gz # which can be measured with featureBits: mv link.tab rbest.link.bed hgLoadBed geoFor1 tLinkMelGal1 rbest.link.bed # Read 17557206 elements of size 5 from rbest.link.bed time featureBits geoFor1 tLinkMelGal1 # 665947116 bases of 1041286029 (63.954%) in intersection # real 1m43.702s hgLoadChain -test -noBin -tIndex geoFor1 \ chainMelGal1RBest geoFor1.melGal1.syn.chain.gz mv link.tab syntenic.link.bed hgLoadBed geoFor1 tLinkMelGal1 syntenic.link.bed # Read 17106461 elements of size 5 from syntenic.link.bed time featureBits geoFor1 tLinkMelGal1 # 653147512 bases of 1041286029 (62.725%) in intersection # real 1m42.611s # cleanup hgsql -e "drop table tLinkMelGal1;" geoFor1 rm chain.tab bed.tab ############################################################################## # lastz galGal4 Chicken (DONE - 2012-02-30 - Hiram) # original alignment to turkey: cd /hive/data/genomes/galGal4/bed/lastzGeoFor1.2012-07-27 cat fb.galGal4.chainGeoFor1Link.txt # 744451026 bases of 1032854810 (72.077%) in intersection # and this swap: mkdir /hive/data/genomes/geoFor1/bed/blastz.galGal4.swap cd /hive/data/genomes/geoFor1/bed/blastz.galGal4.swap time nice -n +19 doBlastzChainNet.pl -verbose=2 \ /hive/data/genomes/galGal4/bed/lastzGeoFor1.2012-07-27/DEF \ -swap -syntenicNet \ -workhorse=hgwdev -smallClusterHub=encodek -bigClusterHub=swarm \ > swap.log 2>&1 & # real 81m0.123s cat fb.geoFor1.chainGalGal4Link.txt # 736380222 bases of 1041286029 (70.718%) in intersection # set sym link to indicate this is the lastz for this genome: cd /hive/data/genomes/geoFor1/bed ln -s blastz.galGal4.swap lastz.galGal4 # lets see what recipBest looks like cd /hive/data/genomes/geoFor1/bed/blastz.galGal4.swap time doRecipBest.pl geoFor1 galGal4 -buildDir=`pwd` -workhorse=hgwdev \ > best.log 2>&1 & # real 46m34.879s cd /hive/data/genomes/geoFor1/bed/blastz.galGal4.swap/axtChain hgLoadChain -test -noBin -tIndex geoFor1 \ chainTestRBest geoFor1.galGal4.rbest.chain.gz # which can be measured with featureBits: mv link.tab rbest.link.bed hgLoadBed geoFor1 tLinkGalGal4 rbest.link.bed # Read 18093500 elements of size 5 from rbest.link.bed time featureBits geoFor1 tLinkGalGal4 # 702254321 bases of 1041286029 (67.441%) in intersection # real 1m44.816s hgLoadChain -test -noBin -tIndex geoFor1 \ chainTestRBest geoFor1.galGal4.syn.chain.gz mv link.tab syntenic.link.bed hgLoadBed geoFor1 tLinkGalGal4 syntenic.link.bed # Read 17708098 elements of size 5 from syntenic.link.bed time featureBits geoFor1 tLinkGalGal4 # 692412176 bases of 1041286029 (66.496%) in intersection # real 1m44.363s # cleanup hgsql -e "drop table tLinkGalGal4;" geoFor1 rm chain.tab bed.tab ############################################################################## # lastz taeGut1 Finch (DONE - 2012-02-31 - Hiram) # original alignment to turkey: cd /hive/data/genomes/taeGut1/bed/lastzGeoFor1.2012-07-28 cat fb.taeGut1.chainGeoFor1Link.txt # 1126123816 bases of 1222864691 (92.089%) in intersection # and for this swap: mkdir /hive/data/genomes/geoFor1/bed/blastz.taeGut1.swap cd /hive/data/genomes/geoFor1/bed/blastz.taeGut1.swap time nice -n +19 doBlastzChainNet.pl -verbose=2 \ /hive/data/genomes/taeGut1/bed/lastzGeoFor1.2012-07-28/DEF \ -swap -syntenicNet \ -workhorse=hgwdev -smallClusterHub=encodek -bigClusterHub=swarm \ > swap.log 2>&1 & # real 102m30.103s cat fb.geoFor1.chainTaeGut1Link.txt # 963520807 bases of 1041286029 (92.532%) in intersection # set sym link to indicate this is the lastz for this genome: cd /hive/data/genomes/geoFor1/bed ln -s blastz.taeGut1.swap lastz.taeGut1 # lets see what recipBest looks like cd /hive/data/genomes/geoFor1/bed/blastz.taeGut1.swap time doRecipBest.pl geoFor1 taeGut1 -buildDir=`pwd` -workhorse=hgwdev \ > best.log 2>&1 & # real 50m20.041s cd /hive/data/genomes/geoFor1/bed/blastz.taeGut1.swap/axtChain hgLoadChain -test -noBin -tIndex geoFor1 \ chainTestRBest geoFor1.taeGut1.rbest.chain.gz # which can be measured with featureBits: mv link.tab rbest.link.bed hgLoadBed geoFor1 tLinkTaeGut1 rbest.link.bed # Read 8455294 elements of size 5 from rbest.link.bed time featureBits geoFor1 tLinkTaeGut1 # 922278077 bases of 1041286029 (88.571%) in intersection # real 1m3.219s hgLoadChain -test -noBin -tIndex geoFor1 \ chainTestRBest geoFor1.taeGut1.syn.chain.gz mv link.tab syntenic.link.bed hgLoadBed geoFor1 tLinkTaeGut1 syntenic.link.bed # Read 8190644 elements of size 5 from syntenic.link.bed time featureBits geoFor1 tLinkTaeGut1 # 914125880 bases of 1041286029 (87.788%) in intersection # real 0m52.942s # cleanup hgsql -e "drop table tLinkTaeGut1;" geoFor1 rm chain.tab bed.tab ############################################################################## ## 7-Way Multiz (DONE - 2012-07-31 - Hiram) ssh hgwdev mkdir /hive/data/genomes/geoFor1/bed/multiz7way cd /hive/data/genomes/geoFor1/bed/multiz7way # from the 63-way in the source tree, select out the 7 used here: /cluster/bin/phast/tree_doctor \ --prune-all-but hg19,mm10,melGal1,melUnd1,galGal4,taeGut1,geoFor1 \ /cluster/home/hiram/kent/src/hg/utils/phyloTrees/63way.nh > 7way.nh # what that looks like: cat 7way.nh # ((hg19:0.148845,mm10:0.356483):0.460153, # (((geoFor1:0.041261,taeGut1:0.034457):0.102066,melUnd1:0.076985):0.224703, # (galGal4:0.032946,melGal1:0.035065):0.106045):0.198405); # rearrange to get geoFor1 on top: cat << '_EOF_' > geoFor1.7way.nh ((((geoFor1:0.041261,taeGut1:0.034457):0.102066,melUnd1:0.076985):0.224703, (galGal4:0.0329457,melGal1:0.035065):0.106045):0.198405, (hg19:0.148845,mm10:0.356483):0.460153); '_EOF_' # << happy emacs # extract species list from that .nh file sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \ geoFor1.7way.nh | xargs echo | sed 's/ //g; s/,/ /g' \ | sed 's/[()]//g; s/,/ /g' | tr '[ ]' '[\n]' > species.list.txt # construct db to name translation list: cat species.list.txt | while read DB do hgsql -N -e "select name,organism from dbDb where name=\"${DB}\";" hgcentraltest done | sed -e "s/\t/->/; s/ /_/g;" | sed -e 's/$/;/' | sed -e 's/\./_/g' \ > db.to.name.txt # construct a common name .nh file: /cluster/bin/phast/tree_doctor --rename \ "`cat db.to.name.txt`" geoFor1.7way.nh | sed -e 's/00*)/)/g; s/00*,/,/g' \ > geoFor1.7way.commonNames.nh # ((((Medium_ground_finch:0.041261,Zebra_finch:0.034457):0.102066, # Budgerigar:0.076985):0.224703, # (Chicken:0.032946,Turkey:0.035065):0.106045):0.198405, # (Human:0.148845,Mouse:0.356483):0.460153); # Use this specification in the phyloGif tool: # http://genome.ucsc.edu/cgi-bin/phyloGif # to obtain a png image for src/hg/htdocs/images/phylo/geoFor1_7way.png /cluster/bin/phast/all_dists geoFor1.7way.nh | grep geoFor1 \ | sed -e "s/geoFor1.//" | sort -k2n > 7way.distances.txt # Use this output to create the table below head 7way.distances.txt # taeGut1 0.075718 # melUnd1 0.220312 # galGal4 0.507021 # melGal1 0.509140 # hg19 1.175433 # mm10 1.383071 cat << '_EOF_' > sizeStats.pl #!/usr/bin/env perl use strict; use warnings; open (FH, "<7way.distances.txt") or die "can not read 7way.distances.txt"; my $count = 0; while (my $line = ) { chomp $line; my ($D, $dist) = split('\s+', $line); my $chain = "chain" . ucfirst($D); my $B="/hive/data/genomes/geoFor1/bed/lastz.$D/fb.geoFor1." . $chain . "Link.txt"; my $chainLinkMeasure = `awk '{print \$5}' ${B} 2> /dev/null | sed -e "s/(//; s/)//"`; chomp $chainLinkMeasure; $chainLinkMeasure = 0.0 if (length($chainLinkMeasure) < 1); $chainLinkMeasure =~ s/\%//; my $swapFile="/hive/data/genomes/${D}/bed/lastz.geoFor1/fb.${D}.chainGeoFor1Link.txt"; my $swapMeasure = "N/A"; if ( -s $swapFile ) { $swapMeasure = `awk '{print \$5}' ${swapFile} 2> /dev/null | sed -e "s/(//; s/)//"`; chomp $swapMeasure; $swapMeasure = 0.0 if (length($swapMeasure) < 1); $swapMeasure =~ s/\%//; } my $orgName= `hgsql -N -e "select organism from dbDb where name='$D';" hgcentraltest`; chomp $orgName; if (length($orgName) < 1) { $orgName="N/A"; } ++$count; printf "# %02d %.4f (%% %06.3f) (%% %06.3f) - %s %s\n", $count, $dist, $chainLinkMeasure, $swapMeasure, $orgName, $D; } close (FH); '_EOF_' # << happy emacs chmod +x ./sizeStats.pl ./sizeStats.pl # # If you can fill in all the numbers in this table, you are ready for # the multiple alignment procedure # featureBits chainLink measures # chainLink # N distance on geoFor1 on other other species # 01 0.0757 (% 92.532) (% 92.089) - Zebra finch taeGut1 # 02 0.2203 (% 80.650) (% 76.598) - Budgerigar melUnd1 # 03 0.5070 (% 70.718) (% 72.077) - Chicken galGal4 # 04 0.5091 (% 67.079) (% 72.625) - Turkey melGal1 # 05 1.1754 (% 08.504) (% 03.503) - Human hg19 # 06 1.3831 (% 07.709) (% 03.543) - Mouse mm10 # None of this concern for distances matters in building the first step, the # maf files. # create species list and stripped down tree for autoMZ sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \ geoFor1.7way.nh | xargs echo | sed 's/ //g; s/,/ /g' > tree.nh sed 's/[()]//g; s/,/ /g' tree.nh > species.list # geoFor1 taeGut1 melUnd1 galGal4 melGal1 hg19 mm10 # bash shell syntax here ... cd /hive/data/genomes/geoFor1/bed/multiz7way export H=/hive/data/genomes/geoFor1/bed mkdir mafLinks for G in `sed -e "s/geoFor1 //" species.list` do mkdir mafLinks/$G if [ -s ${H}/lastz.${G}/mafRBestNet/geoFor1.${G}.rbest.maf.gz ]; then echo "$G - recipBest" echo ln -s ${H}/lastz.$G/mafRBestNet/*.maf.gz ./mafLinks/$G ln -s ${H}/lastz.$G/mafRBestNet/*.maf.gz ./mafLinks/$G else if [ -s ${H}/lastz.${G}/axtChain/geoFor1.${G}.synNet.maf.gz ]; then echo "$G - synNet" echo ln -s ${H}/lastz.$G/axtChain/geoFor1.${G}.synNet.maf.gz ./mafLinks/$G ln -s ${H}/lastz.$G/axtChain/geoFor1.${G}.synNet.maf.gz ./mafLinks/$G else if [ -s ${H}/lastz.${G}/mafNet/geoFor1.${G}.net.maf.gz ]; then echo "$G - mafNet" echo ln -s ${H}/lastz.$G/mafNet/*.maf.gz ./mafLinks/$G ln -s ${H}/lastz.$G/mafNet/*.maf.gz ./mafLinks/$G else echo "missing directory lastz.${G}/*Net" fi fi fi done # would like to use the synNets for the other birds, and recip best # to human and mouse ln -s /hive/data/genomes/geoFor1/bed/lastz.taeGut1/axtChain/geoFor1.taeGut1.synNet.maf.gz ./mafLinks/taeGut1 ln -s /hive/data/genomes/geoFor1/bed/lastz.melGal1/axtChain/geoFor1.melGal1.synNet.maf.gz ./mafLinks/melGal1 ln -s /hive/data/genomes/geoFor1/bed/lastz.galGal4/axtChain/geoFor1.galGal4.synNet.maf.gz ./mafLinks/galGal4 ln -s /hive/data/genomes/geoFor1/bed/lastz.melUnd1/axtChain/geoFor1.melUnd1.synNet.maf.gz ./mafLinks/melUnd1 ln -s /hive/data/genomes/geoFor1/bed/lastz.hg19/mafRBestNet/geoFor1.hg19.rbest.maf.gz ./mafLinks/hg19 ln -s /hive/data/genomes/geoFor1/bed/lastz.mm10/mafRBestNet/geoFor1.mm10.rbest.maf.gz ./mafLinks/mm10 # verify the alignment type is correct: for D in `grep -v geoFor1 /hive/users/hiram/bigWays/geoFor1.7way/ordered.list` do ls -l mafLinks/$D/*.maf.gz | awk '{print $NF}' done # compare to the list at: # http://genomewiki.ucsc.edu/index.php/GeoFor1_Genome_size_statistics # split the maf files into a set of hashed named files # this hash named split keeps the same chr/contig names in the same # named hash file. mkdir /hive/data/genomes/geoFor1/bed/multiz7way/mafSplit cd /hive/data/genomes/geoFor1/bed/multiz7way/mafSplit for D in `sed -e "s/geoFor1 //" ../species.list` do echo "${D}" mkdir $D cd $D echo "mafSplit -byTarget -useHashedName=10 /dev/null . ../../mafLinks/${D}/*.maf.gz" mafSplit -byTarget -useHashedName=8 /dev/null . \ ../../mafLinks/${D}/*.maf.gz cd .. done # construct a list of all possible maf file names. # they do not all exist in each of the species directories find . -type f | wc -l # 1294 find . -type f | grep ".maf$" | xargs -L 1 basename | sort -u > maf.list wc -l maf.list # 256 maf.list mkdir /hive/data/genomes/geoFor1/bed/multiz7way/splitRun cd /hive/data/genomes/geoFor1/bed/multiz7way/splitRun mkdir maf run cd run mkdir penn cp -p /cluster/bin/penn/multiz.2009-01-21/multiz penn cp -p /cluster/bin/penn/multiz.2009-01-21/maf_project penn cp -p /cluster/bin/penn/multiz.2009-01-21/autoMZ penn # set the db and pairs directories here cat > autoMultiz.csh << '_EOF_' #!/bin/csh -ef set db = geoFor1 set c = $1 set result = $2 set run = `/bin/pwd` set tmp = /scratch/tmp/$db/multiz.$c set pairs = /hive/data/genomes/geoFor1/bed/multiz7way/mafSplit /bin/rm -fr $tmp /bin/mkdir -p $tmp /bin/cp -p ../../tree.nh ../../species.list $tmp pushd $tmp > /dev/null foreach s (`/bin/sed -e "s/$db //" species.list`) set in = $pairs/$s/$c set out = $db.$s.sing.maf if (-e $in.gz) then /bin/zcat $in.gz > $out if (! -s $out) then echo "##maf version=1 scoring=autoMZ" > $out endif else if (-e $in) then /bin/ln -s $in $out else echo "##maf version=1 scoring=autoMZ" > $out endif end set path = ($run/penn $path); rehash $run/penn/autoMZ + T=$tmp E=$db "`cat tree.nh`" $db.*.sing.maf $c \ > /dev/null popd > /dev/null /bin/rm -f $result /bin/cp -p $tmp/$c $result /bin/rm -fr $tmp '_EOF_' # << happy emacs chmod +x autoMultiz.csh cat << '_EOF_' > template #LOOP ./autoMultiz.csh $(file1) {check out line+ /hive/data/genomes/geoFor1/bed/multiz7way/splitRun/maf/$(root1).maf} #ENDLOOP '_EOF_' # << happy emacs ln -s ../../mafSplit/maf.list maf.list ssh swarm cd /hive/data/genomes/geoFor1/bed/multiz7way/splitRun/run # the tac reverses the list to get the small jobs first gensub2 maf.list single template stdout | tac > jobList para -ram=8g create jobList # Completed: 256 of 256 jobs # CPU time in finished jobs: 59583s 993.04m 16.55h 0.69d 0.002 y # IO & Wait Time: 6044s 100.74m 1.68h 0.07d 0.000 y # Average job time: 256s 4.27m 0.07h 0.00d # Longest finished job: 2781s 46.35m 0.77h 0.03d # Submission to last job: 2858s 47.63m 0.79h 0.03d # combine into one file (the 1>&2 redirect sends the echo to stderr) cd /hive/data/genomes/geoFor1/bed/multiz7way head -1 splitRun/maf/001.maf > multiz7way.maf for F in splitRun/maf/*.maf do echo "${F}" 1>&2 egrep -v "^#" ${F} done >> multiz7way.maf tail -1 splitRun/maf/001.maf >> multiz7way.maf # -rw-rw-r-- 1 5674242413 Aug 3 08:03 multiz7way.maf # -rw-rw-r-- 1 5670181186 Aug 1 14:41 multiz7way.maf.0 # Load into database ssh hgwdev cd /hive/data/genomes/geoFor1/bed/multiz7way mkdir /gbdb/geoFor1/multiz7way ln -s `pwd`/multiz7way.maf /gbdb/geoFor1/multiz7way cd /scratch/tmp time nice -n +17 hgLoadMaf geoFor1 multiz7way # Loaded 3644645 mafs in 1 files from /gbdb/geoFor1/multiz7way # real 3m40.564s time nice -n +17 hgLoadMafSummary -verbose=2 -minSize=30000 \ -mergeGap=1500 -maxSize=200000 geoFor1 multiz7waySummary \ /gbdb/geoFor1/multiz7way/multiz7way.maf # Created 328873 summary blocks from 10243549 components # and 3644645 mafs from /gbdb/geoFor1/multiz7way/multiz7way.maf # real 4m12.044s wc -l multiz7way*.tab # 3644645 multiz7way.tab # 328873 multiz7waySummary.tab # 3973518 total rm multiz7way*.tab ############################################################################## # GAP ANNOTATE MULTIZ7WAY MAF AND LOAD TABLES (DONE - 2012-08-02 - Hiram) # mafAddIRows has to be run on single chromosome maf files, it does not # function correctly when more than one reference sequence # are in a single file. Need to split of the maf file into individual # maf files mkdir -p /hive/data/genomes/geoFor1/bed/multiz7way/anno/mafSplit cd /hive/data/genomes/geoFor1/bed/multiz7way/anno/mafSplit # XXX this split directory may have been corrupted by a mistaken run of # this command here at a much later date. It had an extra level # of splits added, it did not erase the original level time mafSplit -outDirDepth=1 -byTarget -useFullSequenceName \ /dev/null . ../../multiz7way.maf # real 2m49.680s find . -type f | wc -l # 2944 # check for N.bed files everywhere: cd /hive/data/genomes/geoFor1/bed/multiz7way/anno for DB in `cat ../species.list` do if [ ! -s /hive/data/genomes/${DB}/${DB}.N.bed ]; then echo "MISS: ${DB}" cd /hive/data/genomes/${DB} twoBitInfo -nBed ${DB}.2bit ${DB}.N.bed else echo " OK: ${DB}" fi done cd /hive/data/genomes/geoFor1/bed/multiz7way/anno for DB in `cat ../species.list` do echo "${DB} " ln -s /hive/data/genomes/${DB}/${DB}.N.bed ${DB}.bed echo ${DB}.bed >> nBeds ln -s /hive/data/genomes/${DB}/chrom.sizes ${DB}.len echo ${DB}.len >> sizes done # make sure they all are successful symLinks: ls -ogrtL screen -S geoFor1 # use a screen to control this longish job ssh swarm cd /hive/data/genomes/geoFor1/bed/multiz7way/anno mkdir result for D in `ls mafSplit` do echo mkdir result/${D} mkdir result/${D} done cat << '_EOF_' > template #LOOP mafAddIRows -nBeds=nBeds mafSplit/$(path1) /hive/data/genomes/geoFor1/geoFor1.2bit {check out exists+ result/$(path1)} #ENDLOOP '_EOF_' # << happy emacs find ./mafSplit -type f | sed -e 's#^./mafSplit/##' > maf.list gensub2 maf.list single template jobList # limit jobs to one per node with the ram=8g requirement para -ram=8g create jobList para try ... check ... push ... # Completed: 2944 of 2944 jobs # CPU time in finished jobs: 560s 9.33m 0.16h 0.01d 0.000 y # IO & Wait Time: 8374s 139.57m 2.33h 0.10d 0.000 y # Average job time: 3s 0.05m 0.00h 0.00d # Longest finished job: 14s 0.23m 0.00h 0.00d # Submission to last job: 161s 2.68m 0.04h 0.00d # verify all result files have some content, look for 0 size files: find ./result -type f -size 0 # should see none # or in this manner: find ./result -type f | xargs ls -og | sort -k3nr | tail # combine into one file (the 1>&2 redirect sends the echo to stderr) head -q -n 1 result/0/JH739903.maf > geoFor1.7way.maf find ./result -type f | while read F do echo "${F}" 1>&2 grep -h -v "^#" ${F} done >> geoFor1.7way.maf # these maf files do not have the end marker, this does nothing: # tail -q -n 1 result/0/JH739903.maf >> geoFor1.7way.maf # How about an official end marker: echo "##eof maf" >> geoFor1.7way.maf ls -og # -rw-rw-r-- 1 6488856099 Aug 3 09:02 geoFor1.7way.maf du -hsc geoFor1.7way.maf # 6.1G geoFor1.7way.maf # construct symlinks to get the individual maf files into gbdb: rm /gbdb/geoFor1/multiz7way/multiz7way.maf # remove previous results ln -s `pwd`/geoFor1.7way.maf /gbdb/geoFor1/multiz7way/multiz7way.maf # Load into database cd /scratch/tmp time nice -n +19 hgLoadMaf -pathPrefix=/gbdb/geoFor1/multiz7way \ geoFor1 multiz7way # Loaded 3783877 mafs in 1 files from /gbdb/geoFor1/multiz7way # real 3m9.770s time hgLoadMafSummary -verbose=2 -minSize=30000 \ -mergeGap=1500 -maxSize=200000 geoFor1 multiz7waySummary \ /gbdb/geoFor1/multiz7way/multiz7way.maf # Created 328873 summary blocks from 10243549 components and 3783877 # mafs from /gbdb/geoFor1/multiz7way/multiz7way.maf # real 3m36.650s # -rw-rw-r-- 1 189735418 Aug 3 09:07 multiz7way.tab # -rw-rw-r-- 1 15808340 Aug 3 09:11 multiz7waySummary.tab rm multiz7way*.tab ###################################################################### # MULTIZ7WAY MAF FRAMES (DONE - 2012-08-02 - Hiram) ssh hgwdev mkdir /hive/data/genomes/geoFor1/bed/multiz7way/frames cd /hive/data/genomes/geoFor1/bed/multiz7way/frames # survey all the genomes to find out what kinds of gene tracks they have cat << '_EOF_' > showGenes.csh #!/bin/csh -fe foreach db (`cat ../species.list`) echo -n "${db}: " set tables = `hgsql $db -N -e "show tables like '%Gene%'"` foreach table ($tables) if ($table == "ensGene" || $table == "refGene" || \ $table == "mgcGenes" || $table == "knownGene" || \ $table == "xenoRefGene" ) then set count = `hgsql $db -N -e "select count(*) from $table"` echo -n "${table}: ${count}, " endif end set orgName = `hgsql hgcentraltest -N -e \ "select scientificName from dbDb where name='$db'"` set orgId = `hgsql hg19 -N -e \ "select id from organism where name='$orgName'"` if ($orgId == "") then echo "Mrnas: 0" else set count = `hgsql hg19 -N -e "select count(*) from gbCdnaInfo where organism=$orgId"` echo "Mrnas: ${count}" endif end '_EOF_' # << happy emacs chmod +x ./showGenes.csh time ./showGenes.csh # geoFor1: xenoRefGene: 288791, Mrnas: 1 # taeGut1: ensGene: 19297, refGene: 1176, xenoRefGene: 179552, Mrnas: 98382 # melGal1: ensGene: 17373, xenoRefGene: 203867, Mrnas: 17703 # galGal4: refGene: 5666, xenoRefGene: 192547, Mrnas: 636086 # melUnd1: xenoRefGene: 301402, Mrnas: 26 # hg19: ensGene: 181648, knownGene: 80922, mgcGenes: 31383, refGene: 43528, xenoRefGene: 150519, Mrnas: 10660434 # mm10: ensGene: 95533, knownGene: 59121, mgcGenes: 26768, refGene: 30588, xenoRefGene: 147579, Mrnas: 5216433 # real 1m24.097s # from that summary, use these gene sets: # xenoRefGene - geoFor1 galGal4 melUnd1 # ensGene - taeGut1 melGal1 # knownGene - hg19 mm10 mkdir genes # 1. knownGene: hg19 mma0 for DB in hg19 mm10 do hgsql -N -e "select name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds from knownGene" ${DB} \ | genePredSingleCover stdin stdout | gzip -2c \ > genes/${DB}.gp.gz done # 2. ensGene: for DB in taeGut1 melGal1 do hgsql -N -e "select name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds from ensGene" ${DB} \ | genePredSingleCover stdin stdout | gzip -2c \ > /scratch/tmp/${DB}.tmp.gz mv /scratch/tmp/${DB}.tmp.gz genes/$DB.gp.gz echo "${DB} done" done # 3. xenoRefGene: for DB in geoFor1 galGal4 melUnd1 do hgsql -N -e "select name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds from xenoRefGene" ${DB} \ | genePredSingleCover stdin stdout | gzip -2c \ > /scratch/tmp/${DB}.tmp.gz mv /scratch/tmp/${DB}.tmp.gz genes/$DB.gp.gz echo "${DB} done" done # verify counts for genes are reasonable: for T in genes/*.gz do echo -n "# $T: " zcat $T | cut -f1 | sort | uniq -c | wc -l done # genes/galGal4.gp.gz: 13101 # genes/geoFor1.gp.gz: 14809 # genes/hg19.gp.gz: 20718 # genes/melGal1.gp.gz: 14050 # genes/melUnd1.gp.gz: 15306 # genes/mm10.gp.gz: 20985 # genes/taeGut1.gp.gz: 17343 time (cat ../anno/geoFor1.7way.maf \ | nice -n +19 genePredToMafFrames geoFor1 stdin stdout \ `sed -e "s#\([a-zA-Z0-9]*\)#\1 genes/\1.gp.gz#g" ../species.list` \ | gzip > multiz7wayFrames.bed.gz) # real 3m19.528s # verify there are frames on everything, should be 7 species: zcat multiz7wayFrames.bed.gz | awk '{print $4}' | sort | uniq -c # 140156 galGal4 # 146124 geoFor1 # 183760 hg19 # 171778 melGal1 # 140917 melUnd1 # 203418 mm10 # 149897 taeGut1 # load the resulting file ssh hgwdev cd /hive/data/genomes/geoFor1/bed/multiz7way/frames time hgLoadMafFrames geoFor1 multiz7wayFrames multiz7wayFrames.bed.gz # real 0m9.574s time featureBits -countGaps geoFor1 multiz7wayFrames # 26482387 bases of 1065292181 (2.486%) in intersection # real 0m13.589s # enable the trackDb entries: # frames multiz7wayFrames # irows on # appears to work OK ######################################################################### # Phylogenetic tree from 60-way (DONE - 2012-08-02 - Hiram) mkdir /hive/data/genomes/geoFor1/bed/multiz7way/4d cd /hive/data/genomes/geoFor1/bed/multiz7way/4d # the annotated maf is: ../anno/geoFor1.7way.maf # using xenoRefGene for geoFor1, only transcribed genes hgsql -N -e "select name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds from xenoRefGene where cdsEnd > cdsStart" geoFor1 \ > xenoRefGene.gp genePredSingleCover xenoRefGene.gp stdout | sort > xenoRefGeneNR.gp wc -l xenoRefGeneNR.gp # 15328 xenoRefGeneNR.gp mkdir annoSplit cd annoSplit time mafSplit -verbose=2 -outDirDepth=1 -byTarget -useFullSequenceName \ /dev/null . ../../anno/geoFor1.7way.maf find . -type f | wc -l # 2944 ssh encodek mkdir /hive/data/genomes/geoFor1/bed/multiz7way/4d/run cd /hive/data/genomes/geoFor1/bed/multiz7way/4d/run mkdir ../mfa # newer versions of msa_view have a slightly different operation # the sed of the gp file inserts the reference species in the chr name cat << '_EOF_' > 4d.csh #!/bin/csh -fe set PHASTBIN = /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin set r = "/hive/data/genomes/geoFor1/bed/multiz7way" set c = $1 set infile = $r/4d/annoSplit/$2 set outDir = $r/4d/mfa/$3:h set outfile = $r/4d/mfa/$3 /bin/mkdir -p $outDir cd /scratch/tmp /bin/awk -v C=$c '$2 == C {print}' $r/4d/xenoRefGeneNR.gp | sed -e "s/\t$c\t/\tgeoFor1.$c\t/" > $c.gp set NL=`wc -l $c.gp| gawk '{print $1}'` if ("$NL" != "0") then $PHASTBIN/msa_view --4d --features $c.gp -i MAF $infile -o SS > $c.ss $PHASTBIN/msa_view -i SS --tuple-size 1 $c.ss > $outfile else echo "" > $outfile endif /bin/rm -f $c.gp $c.ss '_EOF_' # << happy emacs chmod +x 4d.csh find ../annoSplit -type f | sed -e "s#../annoSplit/##" > maf.list cat << '_EOF_' > template #LOOP 4d.csh $(file1) $(path1) {check out line+ ../mfa/$(dir1)/$(root1).mfa} #ENDLOOP '_EOF_' # << happy emacs gensub2 maf.list single template jobList para create jobList para try ... check para time # Completed: 2774 of 2944 jobs # Crashed: 170 jobs # CPU time in finished jobs: 818s 13.63m 0.23h 0.01d 0.000 y # IO & Wait Time: 8636s 143.93m 2.40h 0.10d 0.000 y # Average job time: 3s 0.06m 0.00h 0.00d # Longest finished job: 184s 3.07m 0.05h 0.00d # Submission to last job: 624s 10.40m 0.17h 0.01d # not all of them work perfectly. I checked one of the failures, # the SS file had no contents beyond the header # combine mfa files ssh hgwdev cd /hive/data/genomes/geoFor1/bed/multiz7way/4d # remove the broken empty files, size 0 and size 1: find ./mfa -type f -size 0 | xargs rm -f # most interesting, this did not identify files of size 1: # find ./mfa -type f -size 1 find ./mfa -type f | xargs ls -og | awk '$3 == 1' | awk '{print $NF}' \ > empty.list cat empty.list | xargs rm -f #want comma-less species.list /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/msa_view \ --aggregate "`cat ../species.list`" mfa/*/*.mfa | sed s/"> "/">"/ \ > 4d.all.mfa # check they are all in there: grep "^>" 4d.all.mfa # >geoFor1 # >taeGut1 # >melUnd1 # >galGal4 # >melGal1 # >hg19 # >mm10 sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \ ../geoFor1.7way.nh | xargs echo | sed -e 's/ //g' > tree_commas.nh # tree_commas.nh looks like: # ((((geoFor1,taeGut1),(melGal1,galGal4)),melUnd1),(hg19,mm10)) # use phyloFit to create tree model (output is phyloFit.mod) time nice -n +19 \ /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/phyloFit \ --EM --precision MED --msa-format FASTA --subst-mod REV \ --tree tree_commas.nh 4d.all.mfa # real 0m3.872s mv phyloFit.mod all.mod grep TREE all.mod # TREE: # ((((geoFor1:0.0413091,taeGut1:0.0343957):0.101451,melUnd1:0.0837731):0.0304192,(galGal4:0.0329708,melGal1:0.0350493):0.079828):0.218721,(hg19:0.118822,mm10:0.19572):0.218721); # manually constructed two different subset lists: # geoFor1 is on each list since it is the reference and is needed for # everything to work. paste birds.list vertebrate.list # geoFor1 geoFor1 # taeGut1 hg19 # melGal1 mm10 # galGal4 # melUnd1 # on organisms that do not have all species in all files, the file names # need to be filtered. Using this perl script to extract from # the full mfa files, only the subset of species from the four lists: cat << '_EOF_' > filterMfa.pl #!/usr/bin/env perl use strict; use warnings; my $argc = scalar(@ARGV); if ($argc != 1) { printf STDERR "usage: filterMfa.pl \n"; exit 255; } my %dbList; my $file = shift; open (FH, "<$file") or die "can not read $file"; printf STDERR "using list: $file\n"; while (my $db = ) { chomp $db; $dbList{$db} = 1; } close (FH); my $speciesName = $file; $speciesName =~ s/.list//; $file = "4d.all.mfa"; # printf STDERR "processing: %s into %s/%s\n", $file, $speciesName open (FH, "<$file") or die "can not read $file"; open (OF, ">$speciesName.subset.mfa") or die "can not write to $speciesName.subset.mfa"; my $inGroup = 0; while (my $line = ) { if ($line =~ m/^>/) { chomp $line; my $faDbName = $line; $faDbName =~ s/^>//; if (exists($dbList{$faDbName})) { $inGroup = 1; printf OF "> %s\n", $faDbName; } else { $inGroup = 0; } } elsif ($inGroup) { printf OF "%s", $line; } } close (FH); close (OF); '_EOF_' # << happy emacs chmod +x filterMfa.pl # extract each set from the full mfa files, run msa_view on # each subset and construct .nh tree for that subset for N in birds vertebrate do ./filterMfa.pl ${N}.list /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/msa_view \ --aggregate "`cat ${N}.list|xargs echo`" ${N}.subset.mfa \ | sed s/"> "/">"/ > 4d.${N}.mfa /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/tree_doctor \ --no-branchlen --prune-all-but="`cat ${N}.list|xargs echo`" \ tree_commas.nh > tree_commas.${N}.nh done # use phyloFit to create tree model (output is phyloFit.mod) for N in birds vertebrate do time nice -n +19 \ /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/phyloFit \ --EM --precision MED --msa-format FASTA --subst-mod REV \ --tree ./tree_commas.${N}.nh 4d.${N}.mfa mv phyloFit.mod ${N}.mod grep TREE ${N}.mod | sed 's/TREE\:\ //' > ${N}.Nway.nh done grep TREE all.mod | sed -e 's/TREE: //' > all.nh /cluster/bin/phast/all_dists all.nh |grep geoFor1 | sort -k3n \ | sed -e "s/geoFor1.//" # taeGut1 0.075705 # melUnd1 0.226533 # galGal4 0.285978 # melGal1 0.288057 # hg19 0.729443 # mm10 0.806341 ######################################################################### # phastCons 7-way (DONE - 2012-08-03 - Hiram) # split 7way mafs into 10M chunks and generate sufficient statistics # files for # phastCons ssh encodek mkdir -p /hive/data/genomes/geoFor1/bed/multiz7way/cons/SS cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/SS mkdir result done cat << '_EOF_' > mkSS.csh #!/bin/csh -ef set d = $1 set c = $2 set doneDir = done/$d set MAF = /hive/data/genomes/geoFor1/bed/multiz7way/anno/result/$d/$c.maf set WINDOWS = /hive/data/genomes/geoFor1/bed/multiz7way/cons/SS/result/$d/$c set WC = `cat $MAF | wc -l` set NL = `grep "^#" $MAF | wc -l` if ( -s $3 ) then exit 0 endif if ( -s $3.running ) then exit 0 endif /bin/mkdir -p $doneDir /bin/date >> $3.running /bin/rm -fr $WINDOWS /bin/mkdir -p $WINDOWS pushd $WINDOWS > /dev/null if ( $WC != $NL ) then /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/msa_split \ $MAF -i MAF -o SS -r $WINDOWS/$c -w 10000000,0 -I 1000 -B 5000 endif popd > /dev/null /bin/date >> $3 /bin/rm -f $3.running '_EOF_' # << happy emacs chmod +x mkSS.csh cat << '_EOF_' > template #LOOP mkSS.csh $(dir1) $(root1) {check out line+ done/$(dir1)/$(root1)} #ENDLOOP '_EOF_' # << happy emacs # do the easy ones first to see some immediate results find ../../anno/result -type f | sed -e "s#../../anno/result/##" > maf.list gensub2 maf.list single template jobList para create jobList para try ... check ... etc # Completed: 2944 of 2944 jobs # CPU time in finished jobs: 748s 12.47m 0.21h 0.01d 0.000 y # IO & Wait Time: 8590s 143.16m 2.39h 0.10d 0.000 y # Average job time: 3s 0.05m 0.00h 0.00d # Longest finished job: 110s 1.83m 0.03h 0.00d # Submission to last job: 518s 8.63m 0.14h 0.01d # not all of them produce results # there isn't much aligned in these MAF files find ./result -type f | wc -l # 1342 # Run phastCons # This job is I/O intensive in its output files, beware where this # takes place or do not run too many at once. ssh encodek mkdir -p /hive/data/genomes/geoFor1/bed/multiz7way/cons/run.cons cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/run.cons # there are going to be two phastCon runs using this same script. # It triggers off of the current working directory # $cwd:t which is the "grp" in this script. Running: # all and vertebrates cat << '_EOF_' > doPhast.csh #!/bin/csh -fe # ../run.cons/doPhast.csh JH741347.1-6434 0/JH741347 JH741347.1-6434.ss set PHASTBIN = /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin set c = $1 set d = $2 set f = $3 set len = $4 set cov = $5 set rho = $6 set grp = $cwd:t set cons = /hive/data/genomes/geoFor1/bed/multiz7way/cons set tmp = $cons/tmp/$d mkdir -p $tmp set ssSrc = $cons/SS/result set useGrp = "$grp.mod" if (-s $cons/$grp/$grp.non-inf) then ln -s $cons/$grp/$grp.mod $tmp ln -s $cons/$grp/$grp.non-inf $tmp ln -s $ssSrc/$d/$f $tmp else ln -s $ssSrc/$d/$f $tmp ln -s $cons/$grp/$grp.mod $tmp endif pushd $tmp > /dev/null if (-s $grp.non-inf) then $PHASTBIN/phastCons $f $useGrp \ --rho $rho --expected-length $len --target-coverage $cov --quiet \ --not-informative `cat $grp.non-inf` \ --seqname $c --idpref $c --most-conserved $c.bed --score > $c.pp else $PHASTBIN/phastCons $f $useGrp \ --rho $rho --expected-length $len --target-coverage $cov --quiet \ --seqname $c --idpref $c --most-conserved $c.bed --score > $c.pp endif popd > /dev/null mkdir -p pp/$d bed/$d sleep 4 touch pp/$d bed/$d rm -f pp/$d/$c.pp rm -f bed/$d/$c.bed mv $tmp/$c.pp pp/$d mv $tmp/$c.bed bed/$d rm -fr $tmp rmdir --ignore-fail-on-non-empty $cons/tmp/$d:h '_EOF_' # << happy emacs chmod +x doPhast.csh # this template will serve for all runs # root1 == chrom name, file1 == ss file name without .ss suffix cat << '_EOF_' > template #LOOP ../run.cons/doPhast.csh $(root1) $(dir1) $(file1) 45 0.3 0.3 {check out line+ pp/$(dir1)/$(root1).pp} #ENDLOOP '_EOF_' # << happy emacs find ../SS/result -type f | sed -e "s#../SS/result/##" > ss.list wc -l ss.list # 3291 ss.list # Create parasol batch and run it # run for all species cd /hive/data/genomes/geoFor1/bed/multiz7way/cons mkdir -p all cd all # Using the .mod tree cp -p ../../4d/all.mod ./all.mod gensub2 ../run.cons/ss.list single ../run.cons/template jobList para create jobList para try ... check ... para push # Completed: 1342 of 1342 jobs # CPU time in finished jobs: 3368s 56.13m 0.94h 0.04d 0.000 y # IO & Wait Time: 12327s 205.46m 3.42h 0.14d 0.000 y # Average job time: 12s 0.19m 0.00h 0.00d # Longest finished job: 327s 5.45m 0.09h 0.00d # Submission to last job: 1427s 23.78m 0.40h 0.02d # create Most Conserved track cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/all cut -f1 ../../../../chrom.sizes | while read C do ls -d bed/?/${C} 2> /dev/null | while read D do cat ${D}/${C}*.bed done | sort -k1,1 -k2,2n \ | awk '{printf "%s\t%d\t%d\tlod=%d\t%s\n", "'${C}'", $2, $3, $5, $5;}' done > tmpMostConserved.bed /cluster/bin/scripts/lodToBedScore tmpMostConserved.bed > mostConserved.bed # -rw-rw-r-- 1 14833537 Aug 3 13:28 tmpMostConserved.bed # -rw-rw-r-- 1 15224375 Aug 3 13:36 mostConserved.bed # load into database ssh hgwdev cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/all time nice -n +19 hgLoadBed geoFor1 phastConsElements7way mostConserved.bed # Read 427825 elements of size 5 from mostConserved.bed # real 0m6.364s # on human we often try for 5% overall cov, and 70% CDS cov # most bets are off here for that goal, these alignments are too few # and too far between featureBits geoFor1 -enrichment xenoRefGene:cds phastConsElements7way # xenoRefGene:cds 1.907%, phastConsElements7way 5.423%, both 1.370%, # cover 71.82%, enrich 13.24x # --rho 0.3 --expected-length 45 --target-coverage 0.3 # Create merged posterier probability file and wiggle track data files cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/all mkdir downloads # the third sed fixes the chrom names, removing the partition extensions time (find ./pp -type f | sed -e "s#^./##; s#\.# d #g; s#-# m #;" \ | sort -k1,1 -k3,3n | sed -e "s# d #.#g; s# m #-#g;" | xargs cat \ | sed -e 's/\.[0-9][0-9]*-[0-9][0-9]* start/ start/' \ | gzip -c > downloads/phastCons7way.wigFix.gz) # real 14m1.183s # check integrity of data with wigToBigWig time (zcat downloads/phastCons7way.wigFix.gz \ | wigToBigWig -verbose=2 stdin /hive/data/genomes/geoFor1/chrom.sizes \ phastCons7way.bw) > bigWig.log 2>&1 & tail bigWig.log # pid=620: VmPeak: 11016124 kB # real 14m51.983s bigWigInfo phastCons7way.bw # version: 4 # isCompressed: yes # isSwapped: 0 # primaryDataSize: 1,504,251,616 # primaryIndexSize: 33,232,748 # zoomLevels: 10 # chromCount: 1321 # basesCovered: 974,030,471 # mean: 0.097493 # min: 0.000000 # max: 1.000000 # std: 0.228025 # encode those files into wiggle data time (zcat downloads/phastCons7way.wigFix.gz \ | wigEncode stdin phastCons7way.wig phastCons7way.wib) # Converted stdin, upper limit 1.00, lower limit 0.00 # real 6m5.520s du -hsc *.wi? # 929M phastCons7way.wib # 102M phastCons7way.wig # 1.1G total # Load gbdb and database with wiggle. ln -s `pwd`/phastCons7way.wib /gbdb/geoFor1/multiz7way/phastCons7way.wib time nice -n +19 hgLoadWiggle -pathPrefix=/gbdb/geoFor1/multiz7way \ geoFor1 phastCons7way phastCons7way.wig # real 0m37.431s # use to set trackDb.ra entries for wiggle min and max # and verify table is loaded correctly wigTableStats.sh geoFor1 phastCons7way # db.table min max mean count sumData stdDev # viewLimits # geoFor1.phastCons7way 0 1 0.0974934 974030471 9.49615e+07 0.228025 # viewLimits=0:1 # Create histogram to get an overview of all the data time nice -n +19 hgWiggle -doHistogram -db=geoFor1 \ -hBinSize=0.001 -hBinCount=1000 -hMinVal=0.0 -verbose=2 \ phastCons7way > histogram.data 2>&1 # real 3m26.324s # create plot of histogram: cat << '_EOF_' | gnuplot > histo.png set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff set size 1.4, 0.8 set key left box set grid noxtics set grid ytics set title " Medium ground finch geoFor1 Histogram phastCons7way track" set xlabel " phastCons7way score" set ylabel " Relative Frequency" set y2label " Cumulative Relative Frequency (CRF)" set y2range [0:1] set y2tics set yrange [0:0.02] plot "histogram.data" using 2:5 title " RelFreq" with impulses, \ "histogram.data" using 2:7 axes x1y2 title " CRF" with lines '_EOF_' # << happy emacs display histo.png & ######################################################################### ### Create a phastCons data set for Birds # setup birds-only run ssh encode mkdir /hive/data/genomes/geoFor1/bed/multiz7way/cons/birds cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/birds # birds-only: get the birds only tree from the 4d directory cp -p ../../4d/birds.mod ./birds.mod # the others become the non-informative list for phastCons to ignore echo "hg19,mm10" > birds.non-inf gensub2 ../run.cons/ss.list single ../run.cons/template jobList para create jobList para try ... check ... push ... etc. # Completed: 1342 of 1342 jobs # CPU time in finished jobs: 2553s 42.55m 0.71h 0.03d 0.000 y # IO & Wait Time: 8862s 147.70m 2.46h 0.10d 0.000 y # Average job time: 9s 0.14m 0.00h 0.00d # Longest finished job: 35s 0.58m 0.01h 0.00d # Submission to last job: 827s 13.78m 0.23h 0.01d cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/birds # create Most Conserved track cut -f1 ../../../../chrom.sizes | while read C do ls -d bed/?/${C} 2> /dev/null | while read D do cat ${D}/${C}*.bed done | sort -k1,1 -k2,2n \ | awk '{printf "%s\t%d\t%d\tlod=%d\t%s\n", "'${C}'", $2, $3, $5, $5;}' done > tmpMostConserved.bed /cluster/bin/scripts/lodToBedScore tmpMostConserved.bed > mostConserved.bed # load into database ssh hgwdev cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/birds time nice -n +19 hgLoadBed geoFor1 phastConsElements7wayBirds \ mostConserved.bed # Read 330827 elements of size 5 from mostConserved.bed # real 0m2.449s # verify coverage featureBits geoFor1 phastConsElements7wayBirds # 57341045 bases of 1041286029 (5.507%) in intersection # --rho 0.3 --expected-length 45 --target-coverage 0.3 featureBits geoFor1 -enrichment xenoRefGene:cds phastConsElements7wayBirds # xenoRefGene:cds 1.907%, phastConsElements7wayBirds 5.507%, # both 1.154%, cover 60.52%, enrich 10.99x # Create the downloads .pp files, from which the phastCons wiggle data # is calculated # sort by chromName, chromStart so that items are in numerical order # for wigEncode cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/birds mkdir downloads # the third sed fixes the chrom names, removing the partition extensions time (find ./pp -type f | sed -e "s#^./##; s#\.# d #g; s#-# m #;" \ | sort -k1,1 -k3,3n | sed -e "s# d #.#g; s# m #-#g;" | xargs cat \ | sed -e 's/\.[0-9][0-9]*-[0-9][0-9]* start/ start/' \ | gzip -c > downloads/phastCons7way.birds.wigFix.gz) & # real 11m12.204s # encode to bigWig # (warning wigToBigWig process grows to about 36 Gb) # in bash, to avoid the 32 Gb memory limit: time (zcat downloads/phastCons7way.birds.wigFix.gz \ | wigToBigWig stdin ../../../../chrom.sizes phastCons7wayBirds.bw) # real 15m43.842s bigWigInfo phastCons7wayBirds.bw # version: 4 # isCompressed: yes # isSwapped: 0 # primaryDataSize: 1,510,189,951 # primaryIndexSize: 33,232,748 # zoomLevels: 10 # chromCount: 1321 # basesCovered: 974,030,471 # mean: 0.099020 # min: 0.000000 # max: 0.998000 # std: 0.228021 # Create merged posterier probability file and wiggle track data files time (zcat downloads/phastCons7way.birds.wigFix.gz \ | wigEncode stdin phastCons7wayBirds.wig phastCons7wayBirds.wib) # Converted stdin, upper limit 1.00, lower limit 0.00 # real 4m52.174s # if desired to use the bigWig file, loading bigWig table: ln -s `pwd`/phastCons7wayBirds.bw /gbdb/geoFor1/bbi hgsql geoFor1 -e 'drop table if exists phastCons7wayBirds; \ create table phastCons7wayBirds \ (fileName varchar(255) not null); \ insert into phastCons7wayBirds values ("/gbdb/geoFor1/bbi/phastCons7wayBirds.bw");' ## load table with wiggle data ssh hgwdev cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/birds ln -s `pwd`/phastCons7wayBirds.wib \ /gbdb/geoFor1/multiz7way/phastCons7wayBirds.wib time nice -n +19 hgLoadWiggle -pathPrefix=/gbdb/geoFor1/multiz7way geoFor1 \ phastCons7wayBirds phastCons7wayBirds.wig # real 0m13.364s wigTableStats.sh geoFor1 phastCons7wayBirds # db.table min max mean count sumData # geoFor1.phastCons7wayBirds 0 0.998 0.0990203 974030471 9.64488e+07 # stdDev viewLimits # 0.228021 viewLimits=0:0.998 # Create histogram to get an overview of all the data time nice -n +19 hgWiggle -doHistogram \ -hBinSize=0.001 -hBinCount=1000 -hMinVal=0.0 -verbose=2 \ -db=geoFor1 phastCons7wayBirds > histogram.data 2>&1 # real 1m42.507s # create plot of histogram: cat << '_EOF_' | gnuplot > histo.png set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff set size 1.4, 0.8 set key left box set grid noxtics set grid ytics set title " Medium ground finch geoFor1 Histogram phastCons7wayBirds track" set xlabel " phastCons7wayBirds score" set ylabel " Relative Frequency" set y2label " Cumulative Relative Frequency (CRF)" set y2range [0:1] set y2tics set yrange [0:0.02] plot "histogram.data" using 2:5 title " RelFreq" with impulses, \ "histogram.data" using 2:7 axes x1y2 title " CRF" with lines '_EOF_' # << happy emacs display histo.png & ######################################################################### ### Create a phastCons data set for Vertebrate # setup vertebrate-only run ssh encode mkdir /hive/data/genomes/geoFor1/bed/multiz7way/cons/vertebrate cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/vertebrate # vertebrate-only: get the vertebrate only tree from the 4d directory cp -p ../../4d/vertebrate.mod ./vertebrate.mod # the remainders become the non-informative list for phastCons to ignore echo "taeGut1,melUnd1,galGal4,melGal1" > vertebrate.non-inf gensub2 ../run.cons/ss.list single ../run.cons/template jobList para -ram=8g create jobList para try ... check ... push ... etc. # Completed: 1342 of 1342 jobs # CPU time in finished jobs: 2228s 37.14m 0.62h 0.03d 0.000 y # IO & Wait Time: 10120s 168.66m 2.81h 0.12d 0.000 y # Average job time: 9s 0.15m 0.00h 0.00d # Longest finished job: 337s 5.62m 0.09h 0.00d # Submission to last job: 239952s 3999.20m 66.65h 2.78d cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/vertebrate # create Most Conserved track cut -f1 ../../../../chrom.sizes | while read C do ls -d bed/?/${C} 2> /dev/null | while read D do cat ${D}/${C}*.bed done | sort -k1,1 -k2,2n \ | awk '{printf "%s\t%d\t%d\tlod=%d\t%s\n", "'${C}'", $2, $3, $5, $5;}' done > tmpMostConserved.bed # real 5m4.064s /cluster/bin/scripts/lodToBedScore tmpMostConserved.bed > mostConserved.bed # load into database ssh hgwdev cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/vertebrate time nice -n +19 hgLoadBed geoFor1 phastConsElements7wayVertebrate \ mostConserved.bed # Read 188142 elements of size 5 from mostConserved.bed # real 0m1.034s # verify coverage featureBits geoFor1 phastConsElements7wayVertebrate # 27126923 bases of 1041286029 (2.605%) in intersection # --rho 0.3 --expected-length 45 --target-coverage 0.3 featureBits geoFor1 -enrichment xenoRefGene:cds \ phastConsElements7wayVertebrate # xenoRefGene:cds 1.907%, phastConsElements7wayVertebrate 2.605%, # both 1.246%, cover 65.30%, enrich 25.07x # Create the downloads .pp files, from which the phastCons wiggle data # is calculated # sort by chromName, chromStart so that items are in numerical order # for wigEncode cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/vertebrate mkdir downloads # the third sed fixes the chrom names, removing the partition extensions time (find ./pp -type f | sed -e "s#^./##; s#\.# d #g; s#-# m #;" \ | sort -k1,1 -k3,3n | sed -e "s# d #.#g; s# m #-#g;" | xargs cat \ | sed -e 's/\.[0-9][0-9]*-[0-9][0-9]* start/ start/' \ | gzip -c > downloads/phastCons7way.vertebrate.wigFix.gz) & # real 8m46.342s # encode to bigWig time (zcat downloads/phastCons7way.vertebrate.wigFix.gz \ | wigToBigWig stdin ../../../../chrom.sizes phastCons7wayVertebrate.bw) # real 7m5.428s.842s bigWigInfo phastCons7wayVertebrate.bw # version: 4 # isCompressed: yes # isSwapped: 0 # primaryDataSize: 280,864,441 # primaryIndexSize: 33,232,748 # zoomLevels: 10 # chromCount: 1321 # basesCovered: 974,030,471 # mean: 0.032686 # min: 0.000000 # max: 1.000000 # std: 0.157933 # Create merged posterier probability file and wiggle track data files time (zcat downloads/phastCons7way.vertebrate.wigFix.gz \ | wigEncode stdin phastCons7wayVertebrate.wig \ phastCons7wayVertebrate.wib) # Converted stdin, upper limit 1.00, lower limit 0.00 # real 2m57.533s # if desired to use the bigWig file, loading bigWig table: ln -s `pwd`/phastCons7wayVertebrate.bw /gbdb/geoFor1/bbi hgsql geoFor1 -e 'drop table if exists phastCons7wayVertebrate; \ create table phastCons7wayVertebrate \ (fileName varchar(255) not null); \ insert into phastCons7wayVertebrate values ("/gbdb/geoFor1/bbi/phastCons7wayVertebrate.bw");' ## load table with wiggle data ssh hgwdev cd /hive/data/genomes/geoFor1/bed/multiz7way/cons/vertebrate ln -s `pwd`/phastCons7wayVertebrate.wib \ /gbdb/geoFor1/multiz7way/phastCons7wayVertebrate.wib time nice -n +19 hgLoadWiggle -pathPrefix=/gbdb/geoFor1/multiz7way geoFor1 \ phastCons7wayVertebrate phastCons7wayVertebrate.wig # real 0m9.678s wigTableStats.sh geoFor1 phastCons7wayVertebrate # db.table min max mean count sumData # geoFor1.phastCons7wayVertebrate 0 1 0.0326863 974030471 3.18374e+07 # stdDev viewLimits # 0.157933 viewLimits=0:0.822352 # Create histogram to get an overview of all the data time nice -n +19 hgWiggle -doHistogram \ -hBinSize=0.001 -hBinCount=1000 -hMinVal=0.0 -verbose=2 \ -db=geoFor1 phastCons7wayVertebrate > histogram.data 2>&1 # real 1m36.253s # create plot of histogram: cat << '_EOF_' | gnuplot > histo.png set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff set size 1.4, 0.8 set key left box set grid noxtics set grid ytics set title " Medium ground finch geoFor1 Histogram phastCons7wayVertebrate track" set xlabel " phastCons7wayVertebrate score" set ylabel " Relative Frequency" set y2label " Cumulative Relative Frequency (CRF)" set y2range [0:1] set y2tics set yrange [0:0.02] plot "histogram.data" using 2:5 title " RelFreq" with impulses, \ "histogram.data" using 2:7 axes x1y2 title " CRF" with lines '_EOF_' # << happy emacs display histo.png & ############################################################################# # phyloP for 7-way (DONE - 2012-08-03 - Hiram) # run phyloP with score=LRT ssh encodek mkdir /cluster/data/geoFor1/bed/multiz7way/consPhyloP cd /cluster/data/geoFor1/bed/multiz7way/consPhyloP mkdir run.phyloP cd run.phyloP # Adjust model file base composition background and rate matrix to be # representative of the chromosomes in play grep BACKGROUND ../../cons/all/all.mod | awk '{printf "%0.3f\n", $3 + $4}' # 0.485 /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/modFreqs \ ../../cons/all/all.mod 0.485 > all.mod # verify, the BACKGROUND should now be paired up: grep BACK all.mod # BACKGROUND: 0.257500 0.242500 0.242500 0.257500 grep BACKGROUND ../../cons/vertebrate/vertebrate.mod \ | awk '{printf "%0.3f\n", $3 + $4}' # 0.513 grep BACKGROUND ../../cons/birds/birds.mod \ | awk '{printf "%0.3f\n", $3 + $4}' # 0.475 /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/modFreqs \ ../../cons/birds/birds.mod 0.475 > birds.mod # verify, the BACKGROUND should now be paired up: grep BACK birds.mod # BACKGROUND: 0.262500 0.237500 0.237500 0.262500 /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/modFreqs \ ../../cons/vertebrate/vertebrate.mod 0.513 > vertebrate.mod # verify, the BACKGROUND should now be paired up: grep BACK vertebrate.mod # BACKGROUND: 0.243500 0.256500 0.256500 0.243500 0/JH741347/JH741347.1-6434 wigFix/0/JH741347/JH741347.1-6434.wigFix cat << '_EOF_' > doPhyloP.csh #!/bin/csh -fe set PHASTBIN = /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin set f = $1 set d = $f:h set file1 = $f:t set out = $2 set cName = $f:t:r set grp = $cwd:t set cons = /hive/data/genomes/geoFor1/bed/multiz7way/consPhyloP set tmp = $cons/tmp/$grp/$f /bin/rm -fr $tmp /bin/mkdir -p $tmp set ssSrc = "/hive/data/genomes/geoFor1/bed/multiz7way/cons/SS/result/$f" set useGrp = "$grp.mod" /bin/ln -s $cons/run.phyloP/$grp.mod $tmp pushd $tmp > /dev/null $PHASTBIN/phyloP --method LRT --mode CONACC --wig-scores --chrom $cName \ -i SS $useGrp $ssSrc.ss > $file1.wigFix popd > /dev/null /bin/mkdir -p $out:h sleep 4 /bin/touch $out:h /bin/mv $tmp/$file1.wigFix $out /bin/rm -fr $tmp /bin/rmdir --ignore-fail-on-non-empty $cons/tmp/$grp/$d /bin/rmdir --ignore-fail-on-non-empty $cons/tmp/$grp/$d:h /bin/rmdir --ignore-fail-on-non-empty $cons/tmp/$grp /bin/rmdir --ignore-fail-on-non-empty $cons/tmp '_EOF_' # << happy emacs # Create list of chunks find ../../cons/SS/result -type f | grep ".ss$" \ | sed -e "s/.ss$//; s#^../../cons/SS/result/##" > ss.list # make sure the list looks good wc -l ss.list # 5690 ss.list # Create template file # file1 == $chr/$chunk/file name without .ss suffix cat << '_EOF_' > template #LOOP ../run.phyloP/doPhyloP.csh $(path1) {check out line+ wigFix/$(dir1)/$(file1).wigFix} #ENDLOOP '_EOF_' # << happy emacs ###################### Running all species ####################### # setup run for all species mkdir /hive/data/genomes/geoFor1/bed/multiz7way/consPhyloP/all cd /hive/data/genomes/geoFor1/bed/multiz7way/consPhyloP/all rm -fr wigFix mkdir wigFix gensub2 ../run.phyloP/ss.list single ../run.phyloP/template jobList # the -ram=8g will allow only one job per node to slow this down since # it would run too fast otherwise. Either run on one of the small # klusters or use the -ram=8g on the para create para -ram=8g create jobList para try ... check ... etc ... para -maxJob=64 push para time > run.time # Completed: 1342 of 1342 jobs # CPU time in finished jobs: 2832s 47.20m 0.79h 0.03d 0.000 y # IO & Wait Time: 9087s 151.45m 2.52h 0.11d 0.000 y # Average job time: 9s 0.15m 0.00h 0.00d # Longest finished job: 58s 0.97m 0.02h 0.00d # Submission to last job: 629s 10.48m 0.17h 0.01d # make downloads mkdir downloads time (find ./wigFix -type f | sed -e "s#^./##; s#\.# d #g; s#-# m #;" \ | sort -k1,1 -k3,3n | sed -e "s# d #.#g; s# m #-#g;" | xargs cat \ | gzip -c > downloads/phyloP7way.wigFix.gz) & # real 12m29.754s # check integrity of data with wigToBigWig time (zcat downloads/phyloP7way.wigFix.gz \ | wigToBigWig -verbose=2 stdin /hive/data/genomes/geoFor1/chrom.sizes \ phyloP7way.bw) > bigWig.log 2>&1 & egrep "real|VmPeak" bigWig.log # pid=9538: VmPeak: 11016120 kB # real 15m27.107s bigWigInfo phyloP7way.bw # version: 4 # isCompressed: yes # isSwapped: 0 # primaryDataSize: 1,473,509,141 # primaryIndexSize: 33,232,748 # zoomLevels: 10 # chromCount: 1321 # basesCovered: 974,030,471 # mean: -0.001963 # min: -4.711000 # max: 0.934000 # std: 0.621009 # encode those files into wiggle data time (zcat downloads/phyloP7way.wigFix.gz \ | wigEncode stdin phyloP7way.wig phyloP7way.wib) & # Converted stdin, upper limit 0.93, lower limit -4.71 # real 5m2.710s du -hsc *.wi? # 193M phyloP7way.wib # 43M phyloP7way.wig # 236M total # Load gbdb and database with wiggle. ln -s `pwd`/phyloP7way.wib /gbdb/geoFor1/multiz7way/phyloP7way.wib nice hgLoadWiggle -pathPrefix=/gbdb/geoFor1/multiz7way geoFor1 \ phyloP7way phyloP7way.wig # use to set trackDb.ra entries for wiggle min and max # and verify table is loaded correctly wigTableStats.sh geoFor1 phyloP7way # db.table min max mean count sumData # geoFor1.phyloP7way -4.711 0.934 -0.00196263 974030471 -1.91166e+06 # stdDev viewLimits # 0.621009 viewLimits=-3.10701:0.934 # that range is: 0.934+4.711 = 5.645 for hBinSize=0.005645 # Create histogram to get an overview of all the data time nice -n +19 hgWiggle -doHistogram \ -hBinSize=0.005645 -hBinCount=1000 -hMinVal=-4.711 -verbose=2 \ -db=geoFor1 phyloP7way > histogram.data 2>&1 # real 2m8.166s # find out the range for the 2:5 graph grep -v chrom histogram.data | grep "^[0-9]" | ave -col=5 stdin # Q1 0.000005 # median 0.000077 # Q3 0.000478 # average 0.001038 # min 0.000000 # max 0.057772 # count 963 # total 1.000007 # standard deviation 0.003958 # expanding the X axis a bit by limiting the xrange to -2:1 # create plot of histogram: cat << '_EOF_' | gnuplot > histo.png set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff set size 1.4, 0.8 set key left box set grid noxtics set grid ytics set title " Medium ground finch geoFor1 Histogram phyloP7way track" set xlabel " phyloP7way score" set ylabel " Relative Frequency" set y2label " Cumulative Relative Frequency (CRF)" set y2range [0:1] set y2tics set xrange [-2:1] set yrange [0:0.007] plot "histogram.data" using 2:5 title " RelFreq" with impulses, \ "histogram.data" using 2:7 axes x1y2 title " CRF" with lines '_EOF_' # << happy emacs display histo.png & ###################### Running the birds subset ####################### mkdir /hive/data/genomes/geoFor1/bed/multiz7way/consPhyloP/birds cd /hive/data/genomes/geoFor1/bed/multiz7way/consPhyloP/birds rm -fr wigFix mkdir wigFix gensub2 ../run.phyloP/ss.list single ../run.phyloP/template jobList para create jobList para try ... check ... push ... etc ... para time # Completed: 1342 of 1342 jobs # CPU time in finished jobs: 2076s 34.60m 0.58h 0.02d 0.000 y # IO & Wait Time: 8982s 149.70m 2.49h 0.10d 0.000 y # Average job time: 8s 0.14m 0.00h 0.00d # Longest finished job: 24s 0.40m 0.01h 0.00d # Submission to last job: 616s 10.27m 0.17h 0.01d cd /cluster/data/geoFor1/bed/multiz7way/consPhyloP/birds mkdir downloads time (find ./wigFix -type f | sed -e "s#^./##; s#\.# d #g; s#-# m #;" \ | sort -k1,1 -k3,3n | sed -e "s# d #.#g; s# m #-#g;" | xargs cat \ | gzip -c > downloads/phyloP7way.birds.wigFix.gz) & # real 11m58.034s # verify data integrity with the bigWig converter: time (zcat downloads/phyloP7way.birds.wigFix.gz \ | wigToBigWig stdin ../../../../chrom.sizes phyloP7wayBirds.bw) & # real 14m41.297s bigWigInfo phyloP7wayBirds.bw # version: 4 # isCompressed: yes # isSwapped: 0 # primaryDataSize: 1,417,786,751 # primaryIndexSize: 33,232,748 # zoomLevels: 10 # chromCount: 1321 # basesCovered: 974,030,471 # mean: -0.017750 # min: -4.169000 # max: 0.508000 # std: 0.612359 time (zcat downloads/phyloP7way.birds.wigFix.gz \ | wigEncode stdin phyloP7wayBirds.wig phyloP7wayBirds.wib \ > wigEncode.log 2>&1) & # real 12m41.561s # Converted stdin, upper limit 0.51, lower limit -4.17 # if you wanted to use the bigWig data for the track, load bigWig table: ln -s `pwd`/phyloP7wayBirds.bw /gbdb/geoFor1/bbi hgsql geoFor1 -e 'drop table if exists phyloP7wayBirds; \ create table phyloP7wayBirds \ (fileName varchar(255) not null); \ insert into phyloP7wayBirds values ("/gbdb/geoFor1/bbi/phyloP7wayBirds.bw");' # loading the wiggle table: ln -s `pwd`/phyloP7wayBirds.wib /gbdb/geoFor1/multiz7way time nice -n +19 hgLoadWiggle -pathPrefix=/gbdb/geoFor1/multiz7way geoFor1 \ phyloP7wayBirds phyloP7wayBirds.wig # real 0m20.548s wigTableStats.sh geoFor1 phyloP7wayBirds # db.table min max mean count sumData stdDev viewLimits # geoFor1.phyloP7wayBirds -4.169 0.508 -0.0177497 974030471 -1.72888e+07 # stdDev viewLimits # 0.612359 viewLimits=-3.07955:0.508 # that range is: 0.508+4.169 = 4.677 for hBinSize=0.004677 # Create histogram to get an overview of all the data time nice -n +19 hgWiggle -doHistogram \ -hBinSize=0.004677 -hBinCount=1000 -hMinVal=-4.169 -verbose=2 \ -db=geoFor1 phyloP7wayBirds > histogram.data 2>&1 # real 2m44.184s # to see yrange: grep -v "^#" histogram.data | ave -col=5 stdin # max 0.038837 # create plot of histogram: cat << '_EOF_' | gnuplot > histo.png set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff set size 1.4, 0.8 set key left box set grid noxtics set grid ytics set title " Medium ground finch geoFor1 Histogram phyloP7wayBirds track" set xlabel " phyloP7wayBirds score" set ylabel " Relative Frequency" set y2label " Cumulative Relative Frequency (CRF)" set y2range [0:1] set y2tics set yrange [0:0.0388] set xrange [-2.5:0.6] plot "histogram.data" using 2:5 title " RelFreq" with impulses, \ "histogram.data" using 2:7 axes x1y2 title " CRF" with lines '_EOF_' # << happy emacs display histo.png & ###################### Running the vertebrate ####################### mkdir /hive/data/genomes/geoFor1/bed/multiz7way/consPhyloP/vertebrate cd /hive/data/genomes/geoFor1/bed/multiz7way/consPhyloP/vertebrate rm -fr wigFix mkdir wigFix gensub2 ../run.phyloP/ss.list single ../run.phyloP/template jobList para create jobList para try ... check ... push ... etc ... para time # Completed: 1342 of 1342 jobs # CPU time in finished jobs: 1126s 18.77m 0.31h 0.01d 0.000 y # IO & Wait Time: 9076s 151.26m 2.52h 0.11d 0.000 y # Average job time: 8s 0.13m 0.00h 0.00d # Longest finished job: 19s 0.32m 0.01h 0.00d # Submission to last job: 233870s 3897.83m 64.96h 2.71d cd /cluster/data/geoFor1/bed/multiz7way/consPhyloP/vertebrate mkdir downloads time (find ./wigFix -type f | sed -e "s#^./##; s#\.# d #g; s#-# m #;" \ | sort -k1,1 -k3,3n | sed -e "s# d #.#g; s# m #-#g;" | xargs cat \ | gzip -c > downloads/phyloP7way.vertebrate.wigFix.gz) & # real 5m47.255 # verify data integrity with the bigWig encoder: time (zcat downloads/phyloP7way.vertebrate.wigFix.gz \ | wigToBigWig stdin ../../../../chrom.sizes phyloP7wayVertebrate.bw) & # real 9m24.019s # real 15m28.263s bigWigInfo phyloP7wayVertebrate.bw # version: 4 # isCompressed: yes # isSwapped: 0 # primaryDataSize: 367,058,938 # primaryIndexSize: 33,232,748 # zoomLevels: 10 # chromCount: 1321 # basesCovered: 974,030,471 # mean: 0.018190 # min: -1.282000 # max: 0.673000 # std: 0.184936 # if you wanted to use the bigWig data, load bigWig table: ln -s `pwd`/phyloP7wayVertebrate.bw /gbdb/geoFor1/bbi hgsql geoFor1 -e 'drop table if exists phyloP7wayVertebrate; \ create table phyloP7wayVertebrate \ (fileName varchar(255) not null); \ insert into phyloP7wayVertebrate values ("/gbdb/geoFor1/bbi/phyloP7wayVertebrate.bw");' time (zcat downloads/phyloP7way.vertebrate.wigFix.gz \ | wigEncode stdin phyloP7wayVertebrate.wig phyloP7wayVertebrate.wib \ > wigEncode.log 2>&1) & # real 3m20.803s # Converted stdin, upper limit 0.67, lower limit -1.28 # loading the wiggle table: ln -s `pwd`/phyloP7wayVertebrate.wib /gbdb/geoFor1/multiz7way time nice -n +19 hgLoadWiggle -pathPrefix=/gbdb/geoFor1/multiz7way geoFor1 \ phyloP7wayVertebrate phyloP7wayVertebrate.wig # real 0m25.361s wigTableStats.sh geoFor1 phyloP7wayVertebrate # db.table min max mean count sumData stdDev viewLimits # geoFor1.phyloP7wayVertebrate -1.282 0.673 0.0181897 974030471 1.77173e+07 # stdDev viewLimits # 0.184936 viewLimits=-0.906489:0.673 # that range is: 0.673+1.282 = 1.955 for hBinSize=0.001955 # Create histogram to get an overview of all the data time nice -n +19 hgWiggle -doHistogram \ -hBinSize=0.001955 -hBinCount=1000 -hMinVal=-1.282 -verbose=2 \ -db=geoFor1 phyloP7wayVertebrate > histogram.data 2>&1 # real 3m5.929s # to see yrange: grep -v "^#" histogram.data | ave -col=5 stdin # Q1 0.000006 # median 0.000024 # Q3 0.000120 # average 0.002404 # min 0.000000 # max 0.752662 # count 416 # total 0.999998 # standard deviation 0.037246 # create plot of histogram: cat << '_EOF_' | gnuplot > histo.png set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff set size 1.4, 0.8 set key left box set grid noxtics set grid ytics set title " Medium ground finch geoFor1 Histogram phyloP7wayVertebrate track" set xlabel " phyloP7wayVertebrate score" set ylabel " Relative Frequency" set y2label " Cumulative Relative Frequency (CRF)" set y2range [0:1] set y2tics set yrange [0:0.753] set xrange [-0.5:0.5] plot "histogram.data" using 2:5 title " RelFreq" with impulses, \ "histogram.data" using 2:7 axes x1y2 title " CRF" with lines '_EOF_' # << happy emacs display histo.png & ############################################################################# # construct download files for 7-way (DONE - 2012-08-07 - Hiram) mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/multiz7way mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/multiz7way/maf mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/multiz7way/alignments mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phastCons7way mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phastCons7way/birds mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phastCons7way/vertebrate mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phastCons7way/geoFor1.7way.phastCons mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phyloP7way mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phyloP7way/birds mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phyloP7way/vertebrate mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phyloP7way/geoFor1.7way.phyloP7way mkdir /hive/data/genomes/geoFor1/bed/multiz7way/downloads cd /hive/data/genomes/geoFor1/bed/multiz7way/downloads mkdir multiz7way phastCons7way phyloP7way cd multiz7way mkdir maf alignments cd maf time cp -p ../../../anno/geoFor1.7way.maf . # real 9m1.872s time gzip *.maf # real 23m59.878s time md5sum *.maf.gz > md5sum.txt # real 1m55.317s ln -s `pwd`/*.maf.gz `pwd`/md5sum.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/multiz7way/maf cd .. du -hsc maf ../../anno/geoFor1.7way.maf # 1.5G maf # 6.1G ../../anno/geoFor1.7way.maf # 7.6G total ln -s ../../geoFor1.7way.nh . ln -s ../../geoFor1.7way.commonNames.nh . ln -s `pwd`/*.nh \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/multiz7way ##################################################################### cd /hive/data/genomes/geoFor1/bed/multiz7way/downloads/phastCons7way mkdir birds vertebrate geoFor1.7way.phastCons cd birds ln -s ../../../cons/birds/downloads/phastCons7way.birds.wigFix.gz . time md5sum *.gz > md5sum.txt & # real 1m36.881s ln -s `pwd`/*.gz `pwd`/md5sum.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phastCons7way/birds cd ../vertebrate ln -s ../../../cons/vertebrate/downloads/phastCons7way.vertebrate.wigFix.gz . time md5sum *.gz > md5sum.txt & # real real 0m10.940s ln -s `pwd`/*.gz `pwd`/md5sum.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phastCons7way/vertebrate cd ../geoFor1.7way.phastCons ln -s ../../../cons/all/downloads/phastCons7way.wigFix.gz . time md5sum *.gz > md5sum.txt & # real 1m24.173s ln -s `pwd`/*.gz `pwd`/md5sum.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phastCons7way/geoFor1.7way.phastCons cd .. ln -s ../../cons/all/all.mod geoFor1.7way.phastCons.mod ln -s ../../cons/birds/birds.mod geoFor1.7way.phastCons.birds.mod ln -s ../../cons/vertebrate/vertebrate.mod geoFor1.7way.phastCons.vertebrate.mod ln -s ../../cons/all/phastCons7way.bw geoFor1.7way.phastCons.bw ln -s ../../cons/birds/phastCons7wayBirds.bw \ geoFor1.7way.phastCons7wayBirds.bw ln -s ../../cons/vertebrate/phastCons7wayVertebrate.bw \ geoFor1.7way.phastCons7wayVertebrate.bw time md5sum *.mod *.bw > md5sum.txt # real 7m35.579s # obtain the README.txt from mm10/phastCons60way and update for this # situation ln -s `pwd`/*.mod `pwd`/*.bw `pwd`/README.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phastCons7way ##################################################################### cd /hive/data/genomes/geoFor1/bed/multiz7way/downloads/phyloP7way mkdir birds vertebrate geoFor1.7way.phyloP7way cd birds ln -s ../../../consPhyloP/birds/downloads/phyloP7way.birds.wigFix.gz . time md5sum *.gz > md5sum.txt & # real 1m51.913s ln -s `pwd`/*.gz `pwd`/md5sum.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phyloP7way/birds cd ../vertebrate ln -s ../../../consPhyloP/vertebrate/downloads/phyloP7way.vertebrate.wigFix.gz . time md5sum *.gz > md5sum.txt & # real 0m29.125s ln -s `pwd`/*.gz `pwd`/md5sum.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phyloP7way/vertebrate cd ../geoFor1.7way.phyloP7way ln -s ../../../consPhyloP/all/downloads/phyloP7way.wigFix.gz . time md5sum *.gz > md5sum.txt & # real 1m56.253s ln -s `pwd`/*.gz `pwd`/md5sum.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phyloP7way/geoFor1.7way.phyloP7way cd .. ln -s ../../consPhyloP/run.phyloP/all.mod geoFor1.7way.phyloP7way.mod ln -s ../../consPhyloP/run.phyloP/birds.mod ./geoFor1.7way.phyloP.birds.mod ln -s ../../consPhyloP/run.phyloP/vertebrate.mod ./geoFor1.7way.phyloP.vertebrate.mod ln -s ../../consPhyloP/all/phyloP7way.bw geoFor1.7way.phyloP7way.bw ln -s ../../consPhyloP/birds/phyloP7wayBirds.bw \ geoFor1.7way.phyloP7wayBirds.bw ln -s ../../consPhyloP/vertebrate/phyloP7wayVertebrate.bw \ geoFor1.7way.phyloP7wayVertebrate.bw time md5sum *.mod *.bw > md5sum.txt & # real 7m36.111s # obtain the README.txt from mm10/phyloP60way and update for this # situation ln -s `pwd`/*.mod `pwd`/*.bw `pwd`/md5sum.txt `pwd`/README.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/phyloP7way ########################################################################### ## create upstream refGene maf files cd /hive/data/genomes/geoFor1/bed/multiz7way/downloads/multiz7way # bash script #!/bin/sh export geneTbl="xenoRefGene" for S in 1000 2000 5000 do echo "making upstream${S}.maf" featureBits geoFor1 ${geneTbl}:upstream:${S} -fa=/dev/null -bed=stdout \ | perl -wpe 's/_up[^\t]+/\t0/' | sort -k1,1 -k2,2n \ | /cluster/bin/$MACHTYPE/mafFrags geoFor1 multiz7way \ stdin stdout \ -orgs=/hive/data/genomes/geoFor1/bed/multiz7way/species.list \ | gzip -c > upstream${S}.${geneTbl}.maf.gz echo "done upstream${S}.${geneTbl}.maf.gz" done # real 7m22.875s md5sum *.nh *.maf.gz > md5sum.txt # obtain the README.txt from mm10/multiz60way and update for this # situation ln -s `pwd`/*.nh `pwd`/*.maf.gz `pwd`/*.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/geoFor1/multiz7way ############################################################################# # hgPal downloads (DONE - Hiram - 2012-08-10) # FASTA from 7-way for xenoRefGene ssh hgwdev screen -S geoFor1HgPal mkdir /hive/data/genomes/geoFor1/bed/multiz7way/pal cd /hive/data/genomes/geoFor1/bed/multiz7way/pal cat ../species.list | tr '[ ]' '[\n]' > order.list # there are too many scaffolds to allow all the results to end up # in one directory. The procedure here will distribute the results # into ten different directories. This isn't an actual kluster # run because the mafGene command needs database access. # count the scaffolds wc -l ../../../chrom.sizes # 27239 ../../../chrom.sizes # divide by 10 indicates 2724 files per directory, this script # constructs the file names divided into ten directories: export dirCount=0 export count=0 for C in `sort -k2n ../../../chrom.sizes | cut -f1` do echo "${dirCount}/${C}" count=`echo $count | awk '{print $1+1}'` if [ $count -gt 2723 ]; then dirCount=`echo $dirCount | awk '{print $1+1}'` count=0; fi done > file.list # constructing a job list cat << '_EOF_' > template #LOOP ./runOne.csh $(root1) exonNuc/$(path1).fa.gz exonAA/$(path1).fa.gz #ENDLOOP '_EOF_' cat << '_EOF_' > runOne.csh #!/bin/csh -efx set chrom = $1 set resultNuc = $2 set resultAA = $3 set mz=multiz7way set gp=xenoRefGene set db=geoFor1 mkdir -p $resultNuc:h mkdir -p $resultAA:h mafGene -chrom=$chrom -exons -noTrans $db $mz $gp order.list stdout | gzip -c > ${resultNuc} mafGene -chrom=$chrom -exons $db $mz $gp order.list stdout | gzip -c > ${resultAA} '_EOF_' chmod +x runOne.csh gensub2 file.list single template jobList # run 30 of the jobs at a time on hgwdev: cat << '_EOF_' > run30.sh #!/bin/sh export NL=-1 cat jobList | while read L do NL=`echo $NL | awk '{print $1+1}'` if [ "${NL}" -lt 30 ]; then echo "${L} &" ${L} & else NL=0 wait echo "${L} &" ${L} & fi done wait '_EOF_' chmod +x run30.sh ./run30.sh export mz=multiz7way export gp=xenoRefGene export db=geoFor1 time (find ./exonAA -type f | xargs zcat | gzip -c > $gp.$mz.exonAA.fa.gz) time (find ./exonNuc -type f | xargs zcat | gzip -c > $gp.$mz.exonNuc.fa.gz) # real 6m30.702s # check how many genes are used: zcat xenoRefGene.multiz7way.exonAA.fa.gz | grep geoFor1 \ | sed -e "s/_geoFor1.*//" | sort | uniq -c | sort -rn | wc -l # 162565 zcat xenoRefGene.multiz7way.exonNuc.fa.gz | grep geoFor1 \ | sed -e "s/_geoFor1.*//" | sort | uniq -c | sort -rn | wc -l # 162565 # no longer need these files since they have all been packaged up rm -rf exonAA exonNuc # we're only distributing exons at the moment export mz=multiz7way export gp=xenoRefGene export db=geoFor1 export pd=/usr/local/apache/htdocs-hgdownload/goldenPath/$db/$mz/alignments mkdir -p $pd md5sum *.fa.gz > md5sum.txt ln -s `pwd`/$gp.$mz.exonAA.fa.gz $pd/$gp.exonAA.fa.gz ln -s `pwd`/$gp.$mz.exonNuc.fa.gz $pd/$gp.exonNuc.fa.gz ln -s `pwd`/md5sum.txt $pd/ #############################################################################