# for emacs: -*- mode: sh; -*- ############################################################################## # hg38 patch 7 build ############################################################################## ## download sequence, prepare files (DONE - 2016-05-18 - Hiram) ############################################################################## mkdir /hive/data/genomes/hg38/bed/hg38Patch7 cd /hive/data/genomes/hg38/bed/hg38Patch7 mkdir genbank cd genbank time rsync -L -a -P \ rsync://ftp.ncbi.nlm.nih.gov/genomes/genbank/vertebrate_mammalian/Homo_sapiens/all_assembly_versions/GCA_000001405.22_GRCh38.p7/ ./ # sent 20702 bytes received 4076781501 bytes 14180181.58 bytes/sec # total size is 4076211269 speedup is 1.00 # real 4m47.015s # appears to be the entire assembly: faSize GCA_000001405.22_GRCh38.p7_genomic.fna.gz # 3232546710 bases (161368469 N's 3071178241 real 1904219590 upper 1166958651 lower) in 525 sequences in 1 files # Total size: mean 6157231.8 sd 29455433.9 min 970 (KI270394.1) max 248956422 (CM000663.2) median 164536 # %36.10 masked total, %38.00 masked real # so the question is, what is new here compared to what we have in hg38 cd /hive/data/genomes/hg38/bed/hg38Patch7 time faCount genbank/GCA_000001405.22_GRCh38.p7_genomic.fna.gz \ > faCount.GRCH38.p7.txt # real 1m17.032s ~/kent/src/hg/makeDb/doc/hg38/scanAssemblyReport.pl ../../chrom.sizes \ faCount.GRCH38.p7.txt genbank/GCA_000001405.22_GRCh38.p7_assembly_report.txt \ | grep new | sed -e 's/^/# /' # there are 70 new sequences: (verify == 525 (p7) - 455 (hg38 chrom.sizes)) # chr1_KQ031383v1_fix 467143 KQ031383.1 new # chr1_KQ983255v1_alt 278659 KQ983255.1 new # chr1_KN538361v1_fix 305542 KN538361.1 new # chr1_KQ458383v1_alt 349938 KQ458383.1 new # chr1_KN196472v1_fix 186494 KN196472.1 new # chr1_KN196473v1_fix 166200 KN196473.1 new # chr1_KQ458382v1_alt 141019 KQ458382.1 new # chr1_KN196474v1_fix 122022 KN196474.1 new # chr1_KN538360v1_fix 460100 KN538360.1 new # chr1_KQ458384v1_alt 212205 KQ458384.1 new # chr2_KQ031384v1_fix 481245 KQ031384.1 new # chr2_KQ983256v1_alt 535088 KQ983256.1 new # chr2_KN538362v1_fix 208149 KN538362.1 new # chr2_KN538363v1_fix 365499 KN538363.1 new # chr3_KN196475v1_fix 451168 KN196475.1 new # chr3_KN196476v1_fix 305979 KN196476.1 new # chr3_KN538364v1_fix 415308 KN538364.1 new # chr3_KQ031385v1_fix 373699 KQ031385.1 new # chr3_KQ031386v1_fix 165718 KQ031386.1 new # chr4_KQ090013v1_alt 90922 KQ090013.1 new # chr4_KQ090014v1_alt 163749 KQ090014.1 new # chr4_KQ090015v1_alt 236512 KQ090015.1 new # chr4_KQ983257v1_fix 230434 KQ983257.1 new # chr4_KQ983258v1_alt 205407 KQ983258.1 new # chr5_KN196477v1_alt 139087 KN196477.1 new # chr6_KQ090017v1_alt 82315 KQ090017.1 new # chr6_KN196478v1_fix 268330 KN196478.1 new # chr6_KQ031387v1_fix 320750 KQ031387.1 new # chr6_KQ090016v1_fix 245716 KQ090016.1 new # chr7_KQ031388v1_fix 179932 KQ031388.1 new # chr9_KQ090018v1_alt 163882 KQ090018.1 new # chr9_KQ090019v1_alt 134099 KQ090019.1 new # chr9_KN196479v1_fix 330164 KN196479.1 new # chr10_KN538367v1_fix 420164 KN538367.1 new # chr10_KQ090020v1_alt 185507 KQ090020.1 new # chr10_KN196480v1_fix 277797 KN196480.1 new # chr10_KN538365v1_fix 14347 KN538365.1 new # chr10_KN538366v1_fix 85284 KN538366.1 new # chr10_KQ090021v1_fix 264545 KQ090021.1 new # chr11_KQ759759v1_fix 196940 KQ759759.1 new # chr11_KN538368v1_alt 203552 KN538368.1 new # chr11_KN196481v1_fix 108875 KN196481.1 new # chr11_KQ090022v1_fix 181958 KQ090022.1 new # chr12_KQ090023v1_alt 109323 KQ090023.1 new # chr12_KN196482v1_fix 211377 KN196482.1 new # chr12_KN538369v1_fix 541038 KN538369.1 new # chr12_KN538370v1_fix 86533 KN538370.1 new # chr12_KQ759760v1_fix 315610 KQ759760.1 new # chr13_KN538372v1_fix 356766 KN538372.1 new # chr13_KQ090024v1_alt 168146 KQ090024.1 new # chr13_KN196483v1_fix 35455 KN196483.1 new # chr13_KN538373v1_fix 148762 KN538373.1 new # chr13_KQ090025v1_alt 123480 KQ090025.1 new # chr13_KN538371v1_fix 206320 KN538371.1 new # chr15_KN538374v1_fix 4998962 KN538374.1 new # chr15_KQ031389v1_alt 2365364 KQ031389.1 new # chr16_KQ031390v1_alt 169136 KQ031390.1 new # chr16_KQ090026v1_alt 59016 KQ090026.1 new # chr16_KQ090027v1_alt 267463 KQ090027.1 new # chr18_KQ458385v1_alt 205101 KQ458385.1 new # chr18_KQ090028v1_fix 407387 KQ090028.1 new # chr19_KN196484v1_fix 370917 KN196484.1 new # chr19_KQ458386v1_fix 405389 KQ458386.1 new # chr22_KN196485v1_alt 156562 KN196485.1 new # chr22_KN196486v1_alt 153027 KN196486.1 new # chr22_KQ458387v1_alt 155930 KQ458387.1 new # chr22_KQ458388v1_alt 174749 KQ458388.1 new # chr22_KQ759761v1_alt 145162 KQ759761.1 new # chr22_KQ759762v1_fix 101037 KQ759762.1 new # chrY_KN196487v1_fix 101150 KN196487.1 new # how much sequence: ~/kent/src/hg/makeDb/doc/hg38/scanAssemblyReport.pl ../../chrom.sizes \ faCount.GRCH38.p7.txt genbank/GCA_000001405.22_GRCh38.p7_assembly_report.txt \ | grep new | awk '{sum += $2; printf "%d\t%s\n", sum, $0}' | tail # 23260605 chrY_KN196487v1_fix 101150 KN196487.1 new # ^^^^^^^^ total new sequence ~/kent/src/hg/makeDb/doc/hg38/scanAssemblyReport.pl ../../chrom.sizes \ faCount.GRCH38.p7.txt genbank/GCA_000001405.22_GRCh38.p7_assembly_report.txt \ | grep new > new.sequences.list cut -f3 new.sequences.list > extract.new.list awk '{printf "s/%s/%s/; ", $3,$1}' new.sequences.list > genbankToUCSC.sed ~/kent/src/hg/makeDb/doc/hg38/scanAssemblyReport.pl ../../chrom.sizes \ faCount.GRCH38.p7.txt genbank/GCA_000001405.22_GRCh38.p7_assembly_report.txt \ | grep -v new > existing.sequences.list cut -f3 existing.sequences.list > extract.exist.list awk '{printf "s/%s/%s/; ", $3,$1}' new.sequences.list > genbankToUCSC.sed faSomeRecords genbank/GCA_000001405.22_GRCh38.p7_genomic.fna.gz \ extract.new.list stdout | sed -e 's/ .*//;' | \ sed -f genbankToUCSC.sed | gzip -c > hg38Patch7.fa.gz faSomeRecords genbank/GCA_000001405.22_GRCh38.p7_genomic.fna.gz \ extract.exist.list stdout | sed -e 's/ .*//;' | gzip -c > existing.fa.gz # verify same amount of sequence here as hg38: faSize existing.fa.gz # 3209286105 bases (159970322 N's 3049315783 real 1890811945 upper # 1158503838 lower) in 455 sequences in 1 files # Total size: mean 7053376.1 sd 31548372.6 min 970 (KI270394.1) # max 248956422 (CM000663.2) median 161218 # %36.10 masked total, %37.99 masked real # hg38 has different masking head -1 ../../faSize.hg38.2bit.txt # 3209286105 bases (159970322 N's 3049315783 real 1460684798 upper 1588630985 lower) in 455 sequences in 1 files # verify correct amount of patch7 sequence here: faSize hg38Patch7.fa.gz # 23260605 bases (1398147 N's 21862458 real 13407645 upper 8454813 lower) # in 70 sequences in 1 files # Total size: mean 332294.4 sd 632777.9 min 14347 (chr10_KN538365v1_fix) # max 4998962 (chr15_KN538374v1_fix) median 205407 # %36.35 masked total, %38.67 masked real # this is the same total obtained before: # 23260605 chrY_KN196487v1_fix 101150 KN196487.1 new # and both together should equal the original full patch7 sequence time faSize existing.fa.gz hg38Patch7.fa.gz # real 1m19.479s # 3232546710 bases (161368469 N's 3071178241 real 1904219590 upper 1166958651 lower) in 525 sequences in 2 files # Total size: mean 6157231.8 sd 29455433.9 min 970 (KI270394.1) max 248956422 (CM000663.2) median 164536 # %36.10 masked total, %38.00 masked real # same as before: # 3232546710 bases (161368469 N's 3071178241 real 1904219590 upper 1166958651 lower) in 525 sequences in 1 files XXX # construct locations file: ~/kent/src/hg/makeDb/doc/hg38/regionScan.pl extract.new.list \ genbank/GCA_000001405.22_GRCh38.p7_assembly_regions.txt \ > patchLocations.bed # verify correct number of locations, measured 70 before: wc -l patchLocations.bed # 70 patchLocations.bed # separate haplotypes from fix patches for two tracks: grep -v fix patchLocations.bed | sed -e 's/_alt//;' \ | sed -e 's/\tchr.*_/\t/;' | sed -e 's/v/./;' > hg38Patch7Haplotypes.bed grep fix patchLocations.bed | sed -e 's/_fix//;' \ | sed -e 's/\tchr.*_/\t/;' | sed -e 's/v\([0-9]\)$/.\1/;' \ > hg38Patch7Patches.bed # verify nothing lost, should be 70: wc -l hg38*.bed # 28 hg38Patch7Haplotypes.bed # 42 hg38Patch7Patches.bed # 70 total hgLoadBed -type=bed4 hg38 hg38Patch7Haplotypes hg38Patch7Haplotypes.bed # Read 28 elements of size 4 from hg38Patch7Haplotypes.bed hgLoadBed -type=bed4 hg38 hg38Patch7Patches hg38Patch7Patches.bed # Read 42 elements of size 4 from hg38Patch7Patches.bed # construct 2bit file: faToTwoBit hg38Patch7.fa.gz hg38Patch7.unmasked.2bit twoBitInfo hg38Patch7.unmasked.2bit stdout | sort -k2nr > hg38Patch7.chrom.sizes # take a look at that to verify it looks OK: cat hg38Patch7.chrom.sizes | sed -e 's/^/# /;' # chr15_KN538374v1_fix 4998962 # chr15_KQ031389v1_alt 2365364 # chr12_KN538369v1_fix 541038 # chr2_KQ983256v1_alt 535088 # chr2_KQ031384v1_fix 481245 # chr1_KQ031383v1_fix 467143 # chr1_KN538360v1_fix 460100 # chr3_KN196475v1_fix 451168 ... # chrY_KN196487v1_fix 101150 # chr22_KQ759762v1_fix 101037 # chr4_KQ090013v1_alt 90922 # chr12_KN538370v1_fix 86533 # chr10_KN538366v1_fix 85284 # chr6_KQ090017v1_alt 82315 # chr16_KQ090026v1_alt 59016 # chr13_KN196483v1_fix 35455 # chr10_KN538365v1_fix 14347 zcat genbank/GCA_000001405.22_GRCh38.p7_assembly_structure/PATCHES/alt_scaffolds/AGP/alt.scaf.agp.gz \ | sed -f genbankToUCSC.sed > hg38Patch7.agp checkAgpAndFa hg38Patch7.agp hg38Patch7.unmasked.2bit | tail -1 # All AGP and FASTA entries agree - both files are valid XXX - ready to continue - Wed May 18 16:34:01 PDT 2016 ############################################################################# # build hg38Patch7 database (TBD - 2016-01-11 - Hiram) # need this database for netClass operations during the chain/net # construction mkdir /hive/data/genomes/hg38Patch7 cd /hive/data/genomes/hg38Patch7 mkdir /gbdb/hg38Patch7 ln -s /hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit ./ ln -s /hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.agp ./ twoBitInfo hg38Patch7.unmasked.2bit stdout | sort -k2nr > chrom.sizes mkdir -p bed/chromInfo awk '{printf "%s\t%d\t/gbdb/hg38Patch7/hg38Patch7.2bit\n", $1, $2}' \ chrom.sizes > bed/chromInfo/chromInfo.tab hgsql -e 'create database hg38Patch7;' hg38 hgsql hg38Patch7 < $HOME/kent/src/hg/lib/grp.sql hgLoadSqlTab hg38Patch7 chromInfo $HOME/kent/src/hg/lib/chromInfo.sql \ bed/chromInfo/chromInfo.tab hgGoldGapGl -noGl hg38Patch7 hg38Patch7.agp featureBits -or -countGaps hg38Patch7 gold gap # 23260605 bases of 23260605 (100.000%) in intersection hgsql hgcentraltest -e 'INSERT INTO dbDb (name, description, nibPath, organism, defaultPos, active, orderKey, genome, scientificName, htmlPath, hgNearOk, hgPbOk, sourceName, taxId) VALUES ("hg38Patch7", "Mar. 2016", "/gbdb/hg38Patch7", "GRCh38.p7", "chr15_KN538374v1_fix:2000000-2200000", 1, 7751, "GRCh38.p7", "Homo sapiens", "/gbdb/hg38Patch7/html/description.html", 0, 0, "GRCh38 patch 7 Genome Reference Consortium Human Reference 38", 9606); INSERT INTO defaultDb (genome, name) VALUES ("GRCh38.p7", "hg38Patch7"); INSERT INTO genomeClade (genome, clade, priority) VALUES ("GRCh38.p7", "haplotypes", 134);' mkdir html # copy description.html from hg38Patch6/html/description.html # edit to update for Patch7 cp ../../hg38Patch6/html/description.html . mkdir -p /hive/data/genomes/hg38Patch7/bed/gc5Base cd /hive/data/genomes/hg38Patch7/bed/gc5Base hgGcPercent -wigOut -doGaps -file=stdout -win=5 -verbose=0 hg38Patch7 \ ../../hg38Patch7.unmasked.2bit | wigEncode stdin gc5Base.{wig,wib} # Converted stdin, upper limit 100.00, lower limit 0.00 mkdir /gbdb/hg38Patch7/wib ln -s `pwd`/gc5Base.wib /gbdb/hg38Patch7/wib hgLoadWiggle -pathPrefix=/gbdb/hg38Patch7/wib hg38Patch7 gc5Base gc5Base.wig mkdir /hive/data/genomes/hg38Patch7/bed/repeatMasker cd /hive/data/genomes/hg38Patch7/bed/repeatMasker time (doRepeatMasker.pl -bigClusterHub=ku \ -workhorse=hgwdev -dbHost=hgwdev -buildDir=`pwd` hg38Patch7) \ > do.log 2>&1 # real 48m28.216s cat faSize.rmsk.txt # 23260605 bases (1398147 N's 21862458 real 10258301 upper 11604157 lower) # in 70 sequences in 1 files # Total size: mean 332294.4 sd 632777.9 min 14347 (chr10_KN538365v1_fix) # max 4998962 (chr15_KN538374v1_fix) median 205407 # %49.89 masked total, %53.08 masked real mkdir /hive/data/genomes/hg38Patch7/bed/simpleRepeat cd /hive/data/genomes/hg38Patch7/bed/simpleRepeat time (doSimpleRepeat.pl -bigClusterHub=ku -workhorse=hgwdev \ -smallClusterHub=ku -buildDir=`pwd` hg38Patch7) > do.log 2>&1 # real 10m48.418s cat fb.simpleRepeat # 1341203 bases of 21862561 (6.135%) in intersection # the simpleRepeat procedure fails in the cleanup step since there # is no TrfPart directory mkdir /hive/data/genomes/hg38Patch7/bed/windowMasker cd /hive/data/genomes/hg38Patch7/bed/windowMasker time (doWindowMasker.pl -buildDir=`pwd` -workhorse=hgwdev \ -dbHost=hgwdev hg38Patch7) > do.log 2>&1 & time (doWindowMasker.pl -buildDir=`pwd` -workhorse=hgwdev \ -continue=cleanup -dbHost=hgwdev hg38Patch7) > cleanup.log 2>&1 & cd /hive/data/genomes/hg38Patch7 twoBitMask hg38Patch7.rmsk.2bit \ -add bed/simpleRepeat/trfMask.bed hg38Patch7.2bit twoBitToFa hg38Patch7.2bit stdout | faSize stdin # 23260605 bases (1398147 N's 21862458 real 10225906 upper 11636552 lower) # in 70 sequences in 1 files # Total size: mean 332294.4 sd 632777.9 min 14347 (chr10_KN538365v1_fix) # max 4998962 (chr15_KN538374v1_fix) median 205407 # %50.03 masked total, %53.23 masked real # same size as above in original source: # 23260605 bases (1398147 N's 21862458 real 13407645 upper 8454813 lower) ln -s `pwd`/hg38Patch7.2bit /gbdb/hg38Patch7 ######################################################################### # run up idKeys files for ncbiRefSeq (DONE - 2016-05-19 - Hiram) mkdir /hive/data/genomes/hg38Patch7/bed/idKeys cd /hive/data/genomes/hg38Patch7/bed/idKeys time (doIdKeys.pl -buildDir=`pwd` hg38Patch7) > do.log 2>&1 # real 0m19.179s cat hg38Patch7.keySignature.txt # bd7baba9d278a16be04aa308a2336dcd ######################################################################### # ncbiRefSeq Genes (DONE - 2016-05-19 - Hiram) mkdir /hive/data/genomes/hg38Patch7/bed/ncbiRefSeq cd /hive/data/genomes/hg38Patch7/bed/ncbiRefSeq # running step wise as this script is still under development time (~/kent/src/hg/utils/automation/doNcbiRefSeq.pl -buildDir=`pwd` \ -bigClusterHub=ku -dbHost=hgwdev \ -stop=download -fileServer=hgwdev -smallClusterHub=ku -workhorse=hgwdev \ refseq vertebrate_mammalian Homo_sapiens \ GCF_000001405.33_GRCh38.p7 hg38Patch7) > download.log 2>&1 # real 11m43.551s time (~/kent/src/hg/utils/automation/doNcbiRefSeq.pl -buildDir=`pwd` \ -continue=process -bigClusterHub=ku -dbHost=hgwdev \ -stop=process -fileServer=hgwdev -smallClusterHub=ku -workhorse=hgwdev \ refseq vertebrate_mammalian Homo_sapiens \ GCF_000001405.33_GRCh38.p7 hg38Patch7) > process.log 2>&1 # real 12m40.655s time (~/kent/src/hg/utils/automation/doNcbiRefSeq.pl -buildDir=`pwd` \ -continue=load -bigClusterHub=ku -dbHost=hgwdev \ -stop=load -fileServer=hgwdev -smallClusterHub=ku -workhorse=hgwdev \ refseq vertebrate_mammalian Homo_sapiens \ GCF_000001405.33_GRCh38.p7 hg38Patch7) > load.log 2>&1 # real 0m13.277s cat fb.ncbiRefSeq.hg38Patch7.txt # 1727668 bases of 21862561 (7.902%) in intersection ######################################################################### # ucscToINSDC table/track (DONE - 2016-05-19 - Hiram) # the sequence here is working for a 'refseq' assembly with a chrM # situation may be specific depending upon what is available in the assembly mkdir /hive/data/genomes/hg38Patch7/bed/ucscToINSDC cd /hive/data/genomes/hg38Patch7/bed/ucscToINSDC # special script to get the names directly out of the assembly report: ./processReport.pl | sort > ucscToINSDC.txt awk '{printf "%s\t%s\n", $2, $1}' ucscToINSDC.txt | sort > insdcToUcsc.txt # there is no name for chrM/NC_001323.1 sequence, there is no such # sequence with an INSDC name grep -v "^#" /hive/data/genomes/hg38/bed/hg38Patch7/genbank/GCA_000001405.22_GRCh38.p7_assembly_report.txt | cut -f5,7 \ | sed -e 's/na\b/notAvailable/;' | awk '{printf "%s\t%s\n", $1, $2}' \ | sort > insdc.refseq.txt # the sed \b means to match word awk '{printf "%s\t0\t%d\n", $1,$2}' ../../chrom.sizes \ | sort > name.coordinate.tab join insdc.refseq.txt insdcToUcsc.txt | tr '[ ]' '[\t]' | sort -k3 \ | join -2 3 name.coordinate.tab - | tr '[ ]' '[\t]' | cut -f1-3,5 \ > ucscToINSDC.bed # should be same line counts throughout (except for insdc.refseq.txt): wc -l * # 196 insdc.refseq.txt # 130 insdcToUcsc.txt # 130 name.coordinate.tab # 130 ucscToINSDC.bed # 130 ucscToINSDC.txt cut -f1 ucscToINSDC.bed | awk '{print length($0)}' | sort -n | tail -1 # 20 # use the 20 in this sed sed -e "s/21/20/" $HOME/kent/src/hg/lib/ucscToINSDC.sql \ | hgLoadSqlTab hg38Patch7 ucscToINSDC stdin ucscToINSDC.bed checkTableCoords hg38Patch7 # should cover %100 entirely: featureBits -countGaps hg38Patch7 ucscToINSDC # 23260605 bases of 23260605 (100.000%) in intersection join -1 2 <(sort -k2 ucscToINSDC.txt) insdc.refseq.txt | tr '[ ]' '[\t]' \ | sort -k2 | join -2 2 name.coordinate.tab - | tr '[ ]' '[\t]' \ | cut -f1-4 > ucscToRefSeq.bed cut -f1 ucscToRefSeq.bed | awk '{print length($0)}' | sort -n | tail -1 # 20 # use the 20 in this sed sed -e "s/21/20/" $HOME/kent/src/hg/lib/ucscToINSDC.sql \ | sed -e 's/INSDC/RefSeq/g;' > ucscToRefSeq.sql hgLoadSqlTab hg38Patch7 ucscToRefSeq ./ucscToRefSeq.sql ucscToRefSeq.bed checkTableCoords hg38Patch7 -table=ucscToRefSeq # should cover %100 all bases: featureBits -countGaps hg38Patch7 ucscToRefSeq # 23260605 bases of 23260605 (100.000%) in intersection ############################################################################## # cpgIslands on UNMASKED sequence (DONE - 2016-05-19 - Hiram) mkdir /hive/data/genomes/hg38Patch7/bed/cpgIslandsUnmasked cd /hive/data/genomes/hg38Patch7/bed/cpgIslandsUnmasked time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku -buildDir=`pwd` \ -tableName=cpgIslandExtUnmasked \ -maskedSeq=/hive/data/genomes/hg38Patch7/hg38Patch7.unmasked.2bit \ -workhorse=hgwdev -smallClusterHub=ku hg38Patch7) > do.log 2>&1 # real 1m11.915s cat fb.hg38Patch7.cpgIslandExtUnmasked.txt # 340599 bases of 21862561 (1.558%) in intersection ########################################################################## # cpgIslands - (DONE - 2016-05-19 - Hiram) mkdir /hive/data/genomes/hg38Patch7/bed/cpgIslands cd /hive/data/genomes/hg38Patch7/bed/cpgIslands time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku \ -workhorse=hgwdev -smallClusterHub=ku hg38Patch7) > do.log 2>&1 # real 1m12.386s cat fb.hg38Patch7.cpgIslandExt.txt # 327336 bases of 21862561 (1.497%) in intersection ############################################################################## # cytoBandIdeo - (DONE - 2016-05-19 - Hiram) mkdir /hive/data/genomes/hg38Patch7/bed/cytoBand cd /hive/data/genomes/hg38Patch7/bed/cytoBand makeCytoBandIdeo.csh hg38Patch7 ######################################################################### # CREATE MICROSAT TRACK (DONE - 2016-04-19 - Hiram) ssh hgwdev mkdir /hive/data/genomes/hg38Patch7/bed/microsat cd /hive/data/genomes/hg38Patch7/bed/microsat awk '($5==2 || $5==3) && $6 >= 15 && $8 == 100 && $9 == 0 {printf("%s\t%s\t%s\t%dx%s\n", $1, $2, $3, $6, $16);}' \ ../simpleRepeat/simpleRepeat.bed > microsat.bed hgLoadBed hg38Patch7 microsat microsat.bed # Read 262 elements of size 4 from microsat.bed ############################################################################# # lastz alignments to hg38 (DONE - 2016-05-19 - Hiram) ############################################################################# mkdir /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19 cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19 # prepare bits of hg38 sequence to lastz align to the patches, # this is selecting out the specific section of hg38 where the patch # is supposed to match, and setting up lastz parameters rm -fr hg38Bits run.blastz run.blastz/tParts run.blastz/qParts psl \ hg38Bits.lift mkdir -p hg38Bits run.blastz run.blastz/tParts run.blastz/qParts psl cut -f4 ../patchLocations.bed | while read FIX do chr=`grep "${FIX}" ../patchLocations.bed | cut -f1` start=`grep "${FIX}" ../patchLocations.bed | cut -f2` end=`grep "${FIX}" ../patchLocations.bed | cut -f3` bitSize=`echo ${end} ${start} | awk '{printf "%d", $1-$2}'` chrSize=`grep -w "${chr}" ../../../chrom.sizes | cut -f2` fixSize=`grep "${FIX}" ../hg38Patch7.chrom.sizes | cut -f2` printf "${chr}:${start}-${end} vs. ${FIX}:0-${fixSize}\n" 1>&2 twoBitToFa /hive/data/genomes/hg38/hg38.unmasked.2bit:${chr}:${start}-${end} stdout \ | sed -e "s/${chr}:/${FIX}_/g" > hg38Bits/${FIX}.fa fixName=`head -1 hg38Bits/${FIX}.fa | sed -e 's/>//;'` printf "${start}\t${fixName}\t${fixSize}\t${chr}\t${chrSize}\n" 1>&2 printf "${start}\t${fixName}\t${fixSize}\t${chr}\t${chrSize}\n" >> hg38Bits.lift printf "/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/hg38Bits.2bit:${fixName}:0-${bitSize}\n" 1>&2 printf "/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/hg38Bits.2bit:${fixName}:0-${bitSize}\n" > run.blastz/qParts/${fixName}.lst printf "/hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit:${FIX}:0-${fixSize}\n" > run.blastz/tParts/${fixName}.lst printf "/cluster/bin/scripts/blastz-run-ucsc -outFormat psl tParts/${fixName}.lst qParts/${fixName}.lst ../DEF {check out exists ../psl/${fixName}.psl}\n" 1>&2 printf "/cluster/bin/scripts/blastz-run-ucsc -outFormat psl tParts/${fixName}.lst qParts/${fixName}.lst ../DEF {check out exists ../psl/${fixName}.psl}\n" >> run.blastz/jobList done faToTwoBit hg38Bits/*.fa hg38Bits.2bit twoBitInfo hg38Bits.2bit stdout | sort -k2n > hg38Bits.chrom.sizes printf 'BLASTZ=/cluster/bin/penn/lastz-distrib-1.03.66/bin/lastz # human vs human # maximum M allowed with lastz is only 254 BLASTZ_M=254 # lastz does not like the O= and E= lines in the matrix file BLASTZ_Q=/scratch/data/blastz/human_chimp.v2.q BLASTZ_O=600 BLASTZ_E=150 # other parameters from hg18 vs venter1 lastz on advice from Webb BLASTZ_K=10000 BLASTZ_Y=15000 BLASTZ_T=2 # TARGET: Human Hg38Patch7 SEQ1_DIR=/hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit SEQ1_LEN=/hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.chrom.sizes SEQ1_CHUNK=10000000 SEQ1_LAP=10000 SEQ1_IN_CONTIGS=0 SEQ1_LIMIT=1 # QUERY: Human Hg38 #SEQ2_DIR=/scratch/data/hg38/hg38.2bit SEQ2_DIR=/hive/data/genomes/hg38/hg38.unmasked.2bit SEQ2_LEN=/scratch/data/hg38/chrom.sizes SEQ2_CTGDIR=/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/hg38Bits.2bit SEQ2_CTGLEN=/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/hg38Bits.chrom.sizes SEQ2_LIFT=/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/hg38Bits.lift SEQ2_CHUNK=10000000 SEQ2_LAP=0 SEQ2_IN_CONTIGS=0 SEQ2_LIMIT=1 BASE=/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19 TMPDIR=/dev/shm ' > DEF ssh ku cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/run.blastz para create jobList para try ... check ... push ... etc para time > run.time # Completed: 70 of 70 jobs # CPU time in finished jobs: 93s 1.55m 0.03h 0.00d 0.000 y # IO & Wait Time: 190s 3.17m 0.05h 0.00d 0.000 y # Average job time: 4s 0.07m 0.00h 0.00d # Longest finished job: 22s 0.37m 0.01h 0.00d # Submission to last job: 95s 1.58m 0.03h 0.00d # put together the individual results cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19 mkdir pslParts cat psl/chr*.psl | gzip -c > pslParts/hg38Patch7.hg38.psl.gz # constructing a chain from those results mkdir -p /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/axtChain/run cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/axtChain/run time zcat ../../pslParts/hg38Patch7.hg38.psl.gz \ | axtChain -psl -verbose=0 -scoreScheme=/scratch/data/blastz/human_chimp.v2.q -minScore=2000 -linearGap=medium stdin \ /hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit \ /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/hg38Bits.2bit \ stdout \ | chainAntiRepeat /hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit \ /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/hg38Bits.2bit \ stdin hg38Patch7.hg38.preLift.chain # real 0m4.069s liftUp -chainQ hg38Patch7.hg38.lifted.chain \ ../../hg38Bits.lift carry hg38Patch7.hg38.preLift.chain # constructing the net files: cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38.2016-05-19/axtChain chainMergeSort run/hg38Patch7.hg38.lifted.chain \ | gzip -c > hg38Patch7.hg38.all.chain.gz chainSplit chain hg38Patch7.hg38.all.chain.gz # Make nets ("noClass", i.e. without rmsk/class stats which are added later): time chainPreNet hg38Patch7.hg38.all.chain.gz \ ../../hg38Patch7.chrom.sizes \ /hive/data/genomes/hg38/chrom.sizes stdout \ | chainNet stdin -minSpace=1 ../../hg38Patch7.chrom.sizes \ /hive/data/genomes/hg38/chrom.sizes stdout /dev/null \ | netSyntenic stdin noClass.net # real 0m0.912s hgLoadChain -tIndex hg38Patch7 chainHg38 hg38Patch7.hg38.all.chain.gz featureBits hg38Patch7 chainHg38Link # 21000125 bases of 21862561 (96.055%) in intersection netClass -verbose=0 -noAr noClass.net hg38Patch7 hg38 hg38Patch7.hg38.net netFilter -minGap=10 hg38Patch7.hg38.net \ | hgLoadNet -verbose=0 hg38Patch7 netHg38 stdin # Make liftOver chains: netChainSubset -verbose=0 noClass.net hg38Patch7.hg38.all.chain.gz stdout \ | chainStitchId stdin stdout | gzip -c > hg38Patch7.hg38.over.chain.gz # Make axtNet for download: one .axt per hg38Patch7 seq. netSplit noClass.net net cd .. mkdir -p axtNet # beware, tcsh scripting here: foreach f (axtChain/net/*.net) netToAxt $f axtChain/chain/$f:t:r.chain \ ../hg38Patch7.unmasked.2bit \ /hive/data/genomes/hg38/hg38.unmasked.2bit stdout \ | axtSort stdin stdout \ | gzip -c > axtNet/$f:t:r.hg38Patch7.hg38.net.axt.gz end # Make mafNet for multiz: one .maf per hg38Patch7 seq. mkdir -p mafNet # beware, tcsh scripting here: foreach f (axtNet/*.hg38Patch7.hg38.net.axt.gz) axtToMaf -tPrefix=hg38Patch7. -qPrefix=hg38. $f \ ../hg38Patch7.chrom.sizes \ /hive/data/genomes/hg38/chrom.sizes \ stdout \ | gzip -c > mafNet/$f:t:r:r:r:r:r.maf.gz end ############################################################################# # run this same business with hg38 as target, Patch7 sequence as query # (DONE - 2016-05-19,20 - Hiram) mkdir /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19 cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19 rm -fr hg38Bits run.blastz psl hg38Bits.lift mkdir -p hg38Bits run.blastz/tParts run.blastz/qParts psl cut -f4 ../patchLocations.bed | while read FIX do chr=`grep "${FIX}" ../patchLocations.bed | cut -f1` start=`grep "${FIX}" ../patchLocations.bed | cut -f2` end=`grep "${FIX}" ../patchLocations.bed | cut -f3` bitSize=`echo ${end} ${start} | awk '{printf "%d", $1-$2}'` chrSize=`grep -w "${chr}" ../../../chrom.sizes | cut -f2` fixSize=`grep "${FIX}" ../hg38Patch7.chrom.sizes | cut -f2` echo ${chr}:${start}-${end} vs. ${FIX}:0-${fixSize} 1>&2 twoBitToFa /hive/data/genomes/hg38/hg38.unmasked.2bit:${chr}:${start}-${end} stdout \ | sed -e "s/${chr}:/${FIX}_/g" > hg38Bits/${FIX}.fa fixName=`head -1 hg38Bits/${FIX}.fa | sed -e 's/>//;'` echo -e "${start}\t${fixName}\t${fixSize}\t${chr}\t${chrSize}" >> hg38Bits.lift echo -e "/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/hg38Bits.2bit:${fixName}:0-${bitSize}" > run.blastz/tParts/${FIX}.lst echo -e "/hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit:${FIX}:0-${fixSize}" > run.blastz/qParts/${FIX}.lst echo -e "/cluster/bin/scripts/blastz-run-ucsc -outFormat psl tParts/${FIX}.lst qParts/${FIX}.lst ../DEF {check out exists ../psl/${FIX}.psl}" >> run.blastz/jobList done faToTwoBit hg38Bits/*.fa hg38Bits.2bit twoBitInfo hg38Bits.2bit stdout | sort -k2n > hg38Bits.chrom.sizes printf 'BLASTZ=/cluster/bin/penn/lastz-distrib-1.03.66/bin/lastz # human vs human # maximum M allowed with lastz is only 254 BLASTZ_M=254 # lastz does not like the O= and E= lines in the matrix file BLASTZ_Q=/scratch/data/blastz/human_chimp.v2.q BLASTZ_O=600 BLASTZ_E=150 # other parameters from hg18 vs venter1 lastz on advice from Webb BLASTZ_K=10000 BLASTZ_Y=15000 BLASTZ_T=2 # TARGET: Human Hg38 SEQ1_DIR=/hive/data/genomes/hg38/hg38.unmasked.2bit SEQ1_LEN=/scratch/data/hg38/chrom.sizes SEQ1_CTGDIR=/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/hg38Bits.2bit SEQ1_CTGLEN=/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/hg38Bits.chrom.sizes SEQ1_LIFT=/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/hg38Bits.lift SEQ1_CHUNK=10000000 SEQ1_LAP=10000 SEQ1_IN_CONTIGS=0 SEQ1_LIMIT=1 # QUERY: Human Hg38Patch7 SEQ2_DIR=/hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit SEQ2_LEN=/hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.chrom.sizes SEQ2_CHUNK=10000000 SEQ2_LAP=0 SEQ2_IN_CONTIGS=0 SEQ2_LIMIT=1 BASE=/hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19 TMPDIR=/dev/shm ' > DEF ssh ku cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/run.blastz para create jobList para try ... check ... push ... etc para time # Completed: 70 of 70 jobs # CPU time in finished jobs: 89s 1.48m 0.02h 0.00d 0.000 y # IO & Wait Time: 184s 3.07m 0.05h 0.00d 0.000 y # Average job time: 4s 0.07m 0.00h 0.00d # Longest finished job: 22s 0.37m 0.01h 0.00d # Submission to last job: 57s 0.95m 0.02h 0.00d # put together the individual results cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19 mkdir pslParts cat psl/chr*.psl | gzip -c > pslParts/hg38.hg38Patch7.psl.gz # constructing a chain from those results mkdir -p /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/axtChain/run cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/axtChain/run time zcat ../../pslParts/hg38.hg38Patch7.psl.gz \ | axtChain -psl -verbose=0 -scoreScheme=/scratch/data/blastz/human_chimp.v2.q -minScore=2000 -linearGap=medium stdin \ /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/hg38Bits.2bit \ /hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit \ stdout \ | chainAntiRepeat /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/hg38Bits.2bit \ /hive/data/genomes/hg38/bed/hg38Patch7/hg38Patch7.unmasked.2bit \ stdin hg38.hg38Patch7.preLift.chain # real 0m2.716s liftUp hg38.hg38Patch7.lifted.chain \ ../../hg38Bits.lift carry hg38.hg38Patch7.preLift.chain # constructing the net files: cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/axtChain chainMergeSort run/hg38.hg38Patch7.lifted.chain \ | gzip -c > hg38.hg38Patch7.all.chain.gz hgLoadChain -tIndex hg38 chainHg38Patch7 hg38.hg38Patch7.all.chain.gz # Loading 7138 chains into hg38.chainHg38Patch7 featureBits hg38 chainHg38Patch7Link # 20345251 bases of 3049335806 (0.667%) in intersection chainSplit chain hg38.hg38Patch7.all.chain.gz # Make nets ("noClass", i.e. without rmsk/class stats which are added later): time chainPreNet hg38.hg38Patch7.all.chain.gz \ /hive/data/genomes/hg38/chrom.sizes \ ../../hg38Patch7.chrom.sizes stdout \ | chainNet stdin -minSpace=1 /hive/data/genomes/hg38/chrom.sizes \ ../../hg38Patch7.chrom.sizes stdout /dev/null \ | netSyntenic stdin noClass.net # real 0m0.509s time netClass -verbose=0 -noAr noClass.net hg38 hg38Patch7 hg38.hg38Patch7.net # real 0m17.871s netFilter -minGap=10 hg38.hg38Patch7.net \ | hgLoadNet -verbose=0 hg38 netHg38Patch7 stdin # Make liftOver chains: netChainSubset -verbose=0 noClass.net hg38.hg38Patch7.all.chain.gz stdout \ | chainStitchId stdin stdout | gzip -c > hg38.hg38Patch7.over.chain.gz # Make axtNet for download: one .axt per hg38Patch7 seq. netSplit noClass.net net cd .. mkdir -p axtNet # more tcsh here foreach f (axtChain/net/*.net) netToAxt $f axtChain/chain/$f:t:r.chain \ /hive/data/genomes/hg38/hg38.unmasked.2bit \ ../hg38Patch7.unmasked.2bit stdout \ | axtSort stdin stdout \ | gzip -c > axtNet/$f:t:r.hg38.hg38Patch7.net.axt.gz end # Make mafNet for multiz: one .maf per hg38Patch7 seq. mkdir -p mafNet # tcsh loop again foreach f (axtNet/*.hg38.hg38Patch7.net.axt.gz) axtToMaf -tPrefix=hg38. -qPrefix=hg38Patch7. $f \ /hive/data/genomes/hg38/chrom.sizes \ ../hg38Patch7.chrom.sizes \ stdout \ | gzip -c > mafNet/$f:t:r:r:r:r:r.maf.gz end XXX - done to here - Thu May 19 16:19:06 PDT 2016 cd /hive/data/genomes/hg38/bed/hg38Patch7/lastzHg38Patch7.2016-05-19/axtChain mkdir -p queryChains chainSplit -q queryChains hg38.hg38Patch7.all.chain.gz # then run a 'lift over' chain/net on each single one mkdir -p singleLiftOver for F in queryChains/*.chain do C=`basename ${F}` B=`echo ${C} | sed -e "s/.chain//"` chainPreNet -inclHap ${F} /hive/data/genomes/hg38/chrom.sizes \ ../../hg38Patch7.chrom.sizes stdout \ | chainNet -inclHap stdin -minSpace=1 /hive/data/genomes/hg38/chrom.sizes \ ../../hg38Patch7.chrom.sizes singleLiftOver/${B}.raw.net \ /dev/null netSyntenic singleLiftOver/${B}.raw.net singleLiftOver/${B}.noClass.net netFilter -chimpSyn singleLiftOver/${B}.noClass.net > singleLiftOver/${B}.chimpSyn.net netChainSubset -verbose=0 singleLiftOver/${B}.noClass.net \ ${F} stdout \ | chainStitchId stdin stdout > singleLiftOver/${C} printf "${F} -> singleLiftOver/${C}\n" 1>&2 done # put the chains together into one file chainMergeSort singleLiftOver/chr*.chain | gzip -c \ > hg38.hg38Patch7.single.over.chain.gz # -rw-rw-r-- 1 13444 May 20 10:40 hg38.hg38Patch7.single.over.chain.gz # construct psl files from those chains chainToPsl hg38.hg38Patch7.single.over.chain.gz \ /hive/data/genomes/hg38/chrom.sizes \ ../../hg38Patch7.chrom.sizes \ /hive/data/genomes/hg38/hg38.unmasked.2bit \ ../../hg38Patch7.unmasked.2bit \ hg38.hg38Patch7.over.psl # pslCheck reports no errors pslCheck -db=hg38 hg38.hg38Patch7.over.psl # checked: 124 failed: 0 errors: 0 # load this PSL track # this table name prefix altSeqLiftOverPsl is recognized in hgc clicks hgLoadPsl hg38 -table=altSeqLiftOverPslP7 hg38.hg38Patch7.over.psl mkdir /hive/data/genomes/hg38/bed/hg38Patch7/seqExt cd /hive/data/genomes/hg38/bed/hg38Patch7/seqExt twoBitToFa ../hg38Patch7.unmasked.2bit hg38Patch7.fa mkdir -p /gbdb/hg38/hg38Patch7 hg38Patch7 faSplit byname hg38Patch7.fa ./hg38Patch7/ ln -s `pwd`/hg38Patch7/*.fa /gbdb/hg38/hg38Patch7 hgLoadSeq -drop -seqTbl=seqHg38Patch7 -extFileTbl=extHg38Patch7 hg38 \ /gbdb/hg38/hg38Patch7/*.fa ############################################################################# # wrap up scripts and file editing (trackDb, hgdownload) # (DONE - 2016-05-20 - Hiram) # edit ${HOME}/kent/src/hg/makeDb/doc/hg38/patchDescr.pl to reflect this most # recent patch version. cd /hive/data/genomes/hg38/bed/hg38Patch7 # edit patchDescr.pl to update to this patch level ${HOME}/kent/src/hg/makeDb/doc/hg38/patchDescr.pl \ > ${HOME}/kent/src/hg/makeDb/trackDb/human/hg38/hg38Patch7.html mkdir /hive/data/genomes/hg38/bed/hg38Patch7/downloads cd /hive/data/genomes/hg38/bed/hg38Patch7/downloads ln -s ../hg38Patch7.unmasked.2bit hg38Patch7.2bit ln -s ../hg38Patch7.agp ln -s ../hg38Patch7.fa.gz md5sum * > md5sum.txt mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/hg38Patch7 ln -s `pwd`/* /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/hg38Patch7/ # copy README.txt from previous patch and edit as needed - replace the # patch description lines with the output of running cd /hive/data/genomes/hg38/bed/hg38Patch7 ./readmeLines.pl genbank/GCA_000001405.22_GRCh38.p7_assembly_report.txt \ genbank/GCA_000001405.22_GRCh38.p7_assembly_regions.txt \ patchLocations.bed new.sequences.list #############################################################################