690a38c3b39e4f52daab9dcfe00dd0952908d438 hiram Wed Mar 16 17:10:46 2022 -0700 adding knownGene protAA and protNuc fa.gz files per user request refs #29093 diff --git src/hg/makeDb/doc/hg38/multiz100way.txt src/hg/makeDb/doc/hg38/multiz100way.txt index a6fa622..3618147 100644 --- src/hg/makeDb/doc/hg38/multiz100way.txt +++ src/hg/makeDb/doc/hg38/multiz100way.txt @@ -1,2028 +1,2055 @@ ############################################################################# ## 100-Way Multiz (DONE - 2015-04-30 - Hiram) ssh hgwdev mkdir /hive/data/genomes/hg38/bed/multiz100way cd /hive/data/genomes/hg38/bed/multiz100way # the hg38.100way.list was prepared by taking the hg19 100-way list and # updating to most recent assembly versions for a few: # papHam1 -> papAnu2, chlSab1 -> chlSab2, rn5 -> rn6, felCat5 -> felCat8 # bosTau7 -> bosTau8, chrPic1 -> chrPic2, danRer7 -> danRer10 # from the 183-way in the source tree, select out the 100 used here: /cluster/bin/phast/tree_doctor \ --prune-all-but `cat hg38.100way.list | xargs echo | tr '[ ]' '[,]'` \ /cluster/home/hiram/kent/src/hg/utils/phyloTrees/183way.nh \ > hg38.100way.nh # what that looks like: ~/kent/src/hg/utils/phyloTrees/asciiTree.pl hg38.100way.nh | sed -e 's/^/# /;' # ((((((((((((((((((hg38:0.00655, # panTro4:0.00684):0.00422, # gorGor3:0.008964):0.009693, # ponAbe2:0.01894):0.003471, # nomLeu3:0.02227):0.01204, # (((rheMac3:0.004991, # macFas5:0.004991):0.003, # papAnu2:0.008042):0.01061, # chlSab2:0.027000):0.025000):0.021830, # (calJac3:0.03, # ... etc ... # (mayZeb1:0.05, # punNye1:0.050000):0.050000):0.050000):0.100000):0.100000):0.097590, # (oryLat2:0.38197, # xipMac1:0.400000):0.100000):0.015000, # gasAcu1:0.246413):0.045, # gadMor1:0.25):0.22564, # (danRer7:0.430752, # astMex1:0.400000):0.300000):0.143632, # lepOcu1:0.400000):0.326688):0.200000, # petMar2:0.975747); # extract species list from that .nh file sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \ hg38.100way.nh | xargs echo | sed 's/ //g; s/,/ /g' \ | sed 's/[()]//g; s/,/ /g' | tr '[ ]' '[\n]' > species.list.txt # construct db to name translation list: cat species.list.txt | while read DB do hgsql -N -e "select name,organism from dbDb where name=\"${DB}\";" hgcentraltest done | sed -e "s/\t/->/; s/ /_/g;" | sed -e 's/$/;/' | sed -e 's/\./_/g' \ | sed -e 's/-nosed/_nosed/; s/-eating/_eating/;' > db.to.name.txt # construct a common name .nh file: /cluster/bin/phast/tree_doctor --rename \ "`cat db.to.name.txt`" hg38.100way.nh | sed -e 's/00*)/)/g; s/00*,/,/g' \ | $HOME/kent/src/hg/utils/phyloTrees/asciiTree.pl /dev/stdin \ > hg38.100way.commonNames.nh cat hg38.100way.commonNames.nh | sed -e 's/^/# /;' # ((((((((((((((((((Human:0.00655, # Chimp:0.00684):0.00422, # Gorilla:0.008964):0.009693, # Orangutan:0.01894):0.003471, # Gibbon:0.02227):0.01204, # (((Rhesus:0.004991, # Crab_eating_macaque:0.004991):0.003, # Baboon:0.008042):0.01061, # Green_monkey:0.027):0.025):0.02183, # (Marmoset:0.03, # ... etc ... # (Zebra_Mbuna:0.05, # Pundamilia_nyererei:0.05):0.05):0.05):0.1):0.1):0.09759, # (Medaka:0.38197, # Southern_platyfish:0.4):0.1):0.015, # Stickleback:0.246413):0.045, # Atlantic_cod:0.25):0.22564, # (Zebrafish:0.430752, # Mexican_tetra_:0.4):0.3):0.143632, # Spotted_gar:0.4):0.326688):0.2, # Lamprey:0.975747); # 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/hg38_100way.png ~/kent/src/hg/utils/phyloTrees/asciiTree.pl hg38.100way.nh > t.nh ~/kent/src/hg/utils/phyloTrees/scientificNames.sh t.nh \ | $HOME/kent/src/hg/utils/phyloTrees/asciiTree.pl /dev/stdin \ > hg38.100way.scientificNames.nh rm -f t.nh cat hg38.100way.scientificNames.nh | sed -e 's/^/# /;' # ((((((((((((((((((Homo_sapiens:0.00655, # Pan_troglodytes:0.00684):0.00422, # Gorilla_gorilla_gorilla:0.008964):0.009693, # Pongo_pygmaeus_abelii:0.01894):0.003471, # Nomascus_leucogenys:0.02227):0.01204, # (((Macaca_mulatta:0.004991, # Macaca_fascicularis:0.004991):0.003, # Papio_anubis:0.008042):0.01061, # Chlorocebus_sabaeus:0.027):0.025):0.02183, # (Callithrix_jacchus:0.03, # ... etc ... # (Maylandia_zebra:0.05, # Pundamilia_nyererei:0.05):0.05):0.05):0.1):0.1):0.09759, # (Oryzias_latipes:0.38197, # Xiphophorus_maculatus:0.4):0.1):0.015, # Gasterosteus_aculeatus:0.246413):0.045, # Gadus_morhua:0.25):0.22564, # (Danio_rerio:0.430752, # Astyanax_mexicanus:0.4):0.3):0.143632, # Lepisosteus_oculatus:0.4):0.326688):0.2, # Petromyzon_marinus:0.975747); /cluster/bin/phast/all_dists hg38.100way.nh | grep hg38 \ | sed -e "s/hg38.//" | sort -k2n > 100way.distances.txt # Use this output to create the table below cat 100way.distances.txt | sed -e 's/^/# /;' # panTro4 0.013390 # gorGor3 0.019734 # ponAbe2 0.039403 # nomLeu3 0.046204 # macFas5 0.079575 # rheMac3 0.079575 # papAnu2 0.079626 # saiBol1 0.087804 # chlSab2 0.087974 # calJac3 0.107454 # ... etc ... # takFla1 2.018555 # hapBur1 2.020965 # fr3 2.022402 # astMex1 2.037735 # petMar2 2.043162 # danRer7 2.068487 # mayZeb1 2.070965 # punNye1 2.070965 # oryLat2 2.105345 # xipMac1 2.123375 cat << '_EOF_' > sizeStats.pl #!/usr/bin/env perl use strict; use warnings; open (FH, "<100way.distances.txt") or die "can not read 100way.distances.txt"; my $count = 0; while (my $line = ) { chomp $line; my ($D, $dist) = split('\s+', $line); my $chain = "chain" . ucfirst($D); my $B="/hive/data/genomes/hg38/bed/lastz.$D/fb.hg38." . $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.hg38/fb.${D}.chainHg38Link.txt"; my $swapMeasure = 0; 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 hg38 on other - other species # panTro4 01 0.0134 % 93.112 % 95.664 - Chimp panTro4 # gorGor3 02 0.0197 % 88.005 % 91.695 - Gorilla gorGor3 # ponAbe2 03 0.0394 % 89.187 % 89.656 - Orangutan ponAbe2 # nomLeu3 04 0.0462 % 86.379 % 90.470 - Gibbon nomLeu3 # macFas5 05 0.0796 % 85.675 % 87.749 - Crab-eating macaque macFas5 # rheMac3 06 0.0796 % 80.828 % 88.220 - Rhesus rheMac3 # papAnu2 07 0.0796 % 84.179 % 84.502 - Baboon papAnu2 # saiBol1 08 0.0878 % 70.565 % 81.466 - Squirrel monkey saiBol1 # chlSab2 09 0.0880 % 84.393 % 88.264 - Green monkey chlSab2 # calJac3 10 0.1075 % 71.709 % 76.757 - Marmoset calJac3 # oryAfe1 11 0.2469 % 40.563 % 34.102 - Aardvark oryAfe1 # otoGar3 12 0.2703 % 53.196 % 64.899 - Bushbaby otoGar3 # cerSim1 13 0.2851 % 56.633 % 69.232 - White rhinoceros cerSim1 # chrAsi1 14 0.2869 % 33.314 % 29.068 - Cape golden mole chrAsi1 # eptFus1 15 0.3176 % 39.123 % 62.229 - Big brown bat eptFus1 # tupChi1 16 0.3188 % 45.256 % 50.350 - Chinese tree shrew tupChi1 # equCab2 17 0.3195 % 55.459 % 66.600 - Horse equCab2 # camFer1 18 0.3197 % 49.095 % 71.268 - Bactrian camel camFer1 # vicPac2 19 0.3270 % 48.909 % 68.755 - Alpaca vicPac2 # turTru2 20 0.3296 % 49.728 % 61.393 - Dolphin turTru2 # canFam3 21 0.3324 % 50.395 % 60.861 - Dog canFam3 # orcOrc1 22 0.3346 % 50.709 % 64.364 - Killer whale orcOrc1 # speTri2 23 0.3354 % 48.283 % 61.854 - Squirrel speTri2 # pteAle1 24 0.3376 % 48.281 % 71.168 - Black flying-fox pteAle1 # susScr3 25 0.3394 % 44.676 % 57.273 - Pig susScr3 # loxAfr3 26 0.3458 % 45.214 % 42.303 - Elephant loxAfr3 # hetGla2 27 0.3471 % 46.248 % 58.855 - Naked mole-rat hetGla2 # pteVam1 28 0.3510 % 43.924 % 69.545 - Megabat pteVam1 # felCat8 29 0.3586 % 51.684 % 00.000 - Cat felCat8 # ailMel1 30 0.3600 % 48.226 % 61.650 - Panda ailMel1 # musFur1 31 0.3600 % 49.631 % 62.396 - Ferret musFur1 # cavPor3 32 0.3627 % 42.371 % 48.000 - Guinea pig cavPor3 # dasNov3 33 0.3667 % 45.349 % 41.895 - Armadillo dasNov3 # oryCun2 34 0.3769 % 42.911 % 48.360 - Rabbit oryCun2 # panHod1 35 0.3785 % 45.619 % 52.526 - Tibetan antelope panHod1 # lepWed1 36 0.3800 % 51.073 % 65.642 - Weddell seal lepWed1 # odoRosDiv1 37 0.3800 % 52.193 % 64.897 - Pacific walrus odoRosDiv1 # myoDav1 38 0.3876 % 38.713 % 60.286 - David's myotis (bat) myoDav1 # bosTau8 39 0.3885 % 45.975 % 50.440 - Cow bosTau8 # capHir1 40 0.3885 % 45.257 % 52.668 - Domestic goat capHir1 # oviAri3 41 0.3885 % 45.352 % 51.787 - Sheep oviAri3 # myoLuc2 42 0.3902 % 38.591 % 59.729 - Microbat myoLuc2 # eleEdw1 43 0.3936 % 26.839 % 24.181 - Cape elephant shrew eleEdw1 # triMan1 44 0.3966 % 45.346 % 46.939 - Manatee triMan1 # jacJac1 45 0.4100 % 34.153 % 40.300 - Lesser Egyptian jerboa jacJac1 # chiLan1 46 0.4171 % 45.665 % 58.002 - Chinchilla chiLan1 # conCri1 47 0.4441 % 35.911 % 61.017 - Star-nosed mole conCri1 # octDeg1 48 0.4571 % 40.238 % 47.651 - Brush-tailed rat octDeg1 # ochPri3 49 0.4638 % 33.618 % 49.837 - Pika ochPri3 # eriEur2 50 0.4659 % 25.488 % 31.297 - Hedgehog eriEur2 # echTel2 51 0.4928 % 29.310 % 32.753 - Tenrec echTel2 # mm10 52 0.5024 % 31.653 % 35.372 - Mouse mm10 # rn6 53 0.5095 % 31.077 % 34.899 - Rat rn6 # criGri1 54 0.5101 % 33.169 % 42.426 - Chinese hamster criGri1 # mesAur1 55 0.5101 % 30.803 % 43.095 - Golden hamster mesAur1 # micOch1 56 0.5101 % 31.489 % 43.539 - Prairie vole micOch1 # sorAra2 57 0.5137 % 26.905 % 35.760 - Shrew sorAra2 # macEug2 58 0.7620 % 06.378 % 07.213 - Wallaby macEug2 # monDom5 59 0.7657 % 14.370 % 11.996 - Opossum monDom5 # sarHar1 60 0.7900 % 12.754 % 12.925 - Tasmanian devil sarHar1 # apaSpi1 61 0.9281 % 04.611 % 06.320 - Spiny softshell turtle apaSpi1 # cheMyd1 62 0.9281 % 06.296 % 07.960 - Green seaturtle cheMyd1 # chrPic2 63 0.9281 % 06.708 % 08.554 - Painted turtle chrPic2 # pelSin1 64 0.9281 % 05.716 % 07.207 - Chinese softshell turtle pelSin1 # ornAna1 65 0.9531 % 07.769 % 11.784 - Platypus ornAna1 # allMis1 66 1.0332 % 07.561 % 08.665 - American alligator allMis1 # colLiv1 67 1.1532 % 04.712 % 10.220 - Rock pigeon colLiv1 # galGal4 68 1.1655 % 04.696 % 10.888 - Chicken galGal4 # anoCar2 69 1.1751 % 03.593 % 05.222 - Lizard anoCar2 # anaPla1 70 1.1932 % 04.268 % 09.679 - Mallard duck anaPla1 # melGal1 71 1.2100 % 04.328 % 10.927 - Turkey melGal1 # falChe1 72 1.2432 % 04.828 % 09.878 - Saker falcon falChe1 # falPer1 73 1.2432 % 04.926 % 10.027 - Peregrine falcon falPer1 # amaVit1 74 1.2539 % 04.186 % 09.480 - Parrot amaVit1 # araMac1 75 1.2539 % 03.787 % 09.039 - Scarlet Macaw araMac1 # pseHum1 76 1.2629 % 05.146 % 11.937 - Tibetan ground jay pseHum1 # melUnd1 77 1.2649 % 04.474 % 09.862 - Budgerigar melUnd1 # ficAlb2 78 1.2950 % 04.822 % 10.748 - Collared flycatcher ficAlb2 # zonAlb1 79 1.3144 % 04.494 % 10.570 - White-throated sparrow zonAlb1 # geoFor1 80 1.3212 % 04.445 % 10.366 - Medium ground finch geoFor1 # taeGut2 81 1.3250 % 05.893 % 12.356 - Zebra finch taeGut2 # lepOcu1 82 1.5941 % 02.511 % 06.634 - Spotted gar lepOcu1 # xenTro7 83 1.6340 % 03.811 % 07.967 - X. tropicalis xenTro7 # latCha1 84 1.7340 % 02.873 % 03.449 - Coelacanth latCha1 # gadMor1 85 1.8134 % 01.660 % 06.911 - Atlantic cod gadMor1 # gasAcu1 86 1.8548 % 02.080 % 11.956 - Stickleback gasAcu1 # oreNil2 87 1.9210 % 01.868 % 06.241 - Nile tilapia oreNil2 # tetNig2 88 1.9427 % 01.743 % 14.323 - Tetraodon tetNig2 # neoBri1 89 1.9710 % 01.767 % 06.840 - Princess of Burundi neoBri1 # takFla1 90 2.0186 % 01.548 % 11.317 - Yellowbelly pufferfish takFla1 # hapBur1 91 2.0210 % 01.783 % 06.861 - Burton's mouthbreeder hapBur1 # fr3 92 2.0224 % 01.784 % 12.394 - Fugu fr3 # astMex1 93 2.0377 % 02.225 % 06.365 - Mexican tetra (cavefish) astMex1 # petMar2 94 2.0432 % 01.265 % 03.960 - Lamprey petMar2 # danRer10 95 2.0685 % 03.357 % 07.110 - Zebrafish danRer10 # mayZeb1 96 2.0710 % 01.805 % 06.780 - Zebra Mbuna mayZeb1 # punNye1 97 2.0710 % 01.787 % 06.864 - Pundamilia nyererei punNye1 # oryLat2 98 2.1053 % 02.002 % 06.853 - Medaka oryLat2 # xipMac1 99 2.1234 % 01.898 % 07.435 - Southern platyfish xipMac1 # 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' \ hg38.100way.nh | xargs echo | sed 's/ //g; s/,/ /g' > tree.nh sed 's/[()]//g; s/,/ /g' tree.nh > species.list # hg38 panTro4 gorGor3 ponAbe2 nomLeu3 rheMac3 macFas5 ... etc ... # oryLat2 xipMac1 gasAcu1 gadMor1 danRer10 astMex1 lepOcu1 petMar2 # scan sizes and N50s to determine quality of assemblies: cat << '_EOF_' > checkN50.sh #!/bin/sh grep -v hg38 hg38.100way.list | while read D do export bases=0 export Ns=0 export ratio=0 export dist=`grep -w "${D}" 100way.distances.txt | awk '{print $2}'` faSize="/hive/data/genomes/${D}/faSize.${D}.2bit.txt" if [ -s "${faSize}" ]; then bases=`grep -w bases "${faSize}" | awk '{print $1}'` Ns=`grep -w bases "${faSize}" | awk '{print $3}' | sed -e 's/(//;'` ratio=`echo $Ns $bases | awk '{printf "%.2f", 100.0*$1/$2}'` else echo "missing ${faSize}" 1>&2 fi if [ -s /hive/data/inside/lastzRuns/n50Data/${D}.n50.txt ]; then echo -n -e "${D}\t$dist\t$bases\t$Ns\t$ratio\t" cut -f4,5 /hive/data/inside/lastzRuns/n50Data/${D}.n50.txt else echo "missing /hive/data/inside/lastzRuns/n50Data/${D}.n50.txt" 1>&2 fi done '_EOF_' # << happy emacs chmod +x checkN50.sh # distance greater than 0.7 can be 'netOnly' subset ./checkN50.sh | awk '$2 > 0.7' | cut -f1 | sort > netOnly.list # percent Ns > %20 or contig count over 2000 or N50 less than 1,000,000 # can be the 'recipBest" subset ./checkN50.sh | awk '$5 > 20 || $6 > 2000 || $7 < 1000000' \ | awk '$2 <= 0.70' | cut -f1 | sort > recipBest.list # the rest can be syntenic net ./checkN50.sh | awk '$5 <= 20 && $6 <= 2000 && $7 >= 1000000' \ | awk '$2 <= 0.70' | cut -f1 | sort > synNet.list # should have 99: wc -l recipBest.list netOnly.list synNet.list # 8 recipBest.list # 42 netOnly.list # 49 synNet.list # 99 total cat recipBest.list netOnly.list synNet.list | sort -u | wc -l # 99 # bash shell syntax here ... cd /hive/data/genomes/hg38/bed/multiz100way export H=/hive/data/genomes/hg38/bed mkdir mafLinks mafSplit -byTarget -useFullSequenceName /dev/null ./hg38_ hg38.ailMel1.synNet.maf.gz # good assemblies can use syntenic net: cat synNet.list | while read G do mkdir mafLinks/$G echo ln -s ${H}/lastz.$G/axtChain/hg38.${G}.synNet.maf.gz ./mafLinks/$G ln -s ${H}/lastz.$G/axtChain/hg38.${G}.synNet.maf.gz ./mafLinks/$G done # poor assemblies using recip best net: cat recipBest.list | while read G do mkdir mafLinks/$G echo ln -s ${H}/lastz.$G/mafRBestNet/hg38.${G}.rbest.maf.gz ./mafLinks/$G ln -s ${H}/lastz.$G/mafRBestNet/hg38.${G}.rbest.maf.gz ./mafLinks/$G done # distant assemblies using ordinary 'net' maf: cat netOnly.list | while read G do mkdir mafLinks/$G echo ln -s ${H}/lastz.$G/mafNet/*.maf.gz ./mafLinks/$G ln -s ${H}/lastz.$G/mafNet/*.maf.gz ./mafLinks/$G done # verify the alignment type file is correct: grep -v hg38 species.list.txt | while read D do ls -og mafLinks/$D/*.maf.gz 2> /dev/null | awk '{print $NF}' done | awk -F'.' '{print $(NF-2)}' | sort | uniq -c # 42 net # 8 rbest # 49 synNet # need to split these things up into smaller pieces for # efficient kluster run. mkdir /hive/data/genomes/hg38/bed/multiz100way/mafSplit cd /hive/data/genomes/hg38/bed/multiz100way/mafSplit # mafSplitPos splits on gaps or repeat areas that will not have # any chains, approx 5 Mbp intervals, gaps at least 10,000 mafSplitPos -minGap=10000 hg38 5 stdout | sort -u \ | sort -k1,1 -k2,2n > mafSplit.bed # There is a splitRegions.pl script here(copied from previous hg19 46way) # that can create a custom track from this mafSplit.bed file. # Take a look at that in the browser and see if it looks OK, # check the number of sections on each chrom to verify none are # too large. Despite the claim above, it does appear that some # areas are split where actual chains exist. ./splitRegions.pl mafSplit.bed > splitRegions.ct # to see the sizes of the regions: grep "^chr" splitRegions.ct | awk '{print $3-$2,$0}' | sort -rn | less # run a kluster job to split them all ssh ku cd /hive/data/genomes/hg38/bed/multiz100way/mafSplit cat << '_EOF_' > runOne #!/bin/csh -ef set G = $1 set M = $2 mkdir -p $G pushd $G > /dev/null if ( -s hg38_${M}.00.maf ) then /bin/rm -f hg38_${M}.*.maf endif /cluster/bin/x86_64/mafSplit ../mafSplit.bed hg38_ ../../mafLinks/${G}/${M}.maf.gz /bin/gzip hg38_*.maf popd > /dev/null '_EOF_' # << happy emacs chmod +x runOne cat << '_EOF_' > template #LOOP runOne $(dir1) $(file1) {check out exists+ $(dir1)/hg38_chr1.00.maf.gz} #ENDLOOP '_EOF_' # << happy emacs find ../mafLinks -type l | awk -F'/' '{printf "%s/%s\n", $3,$4}' \ | sed -e 's/.maf.gz//;' > maf.list gensub2 maf.list single template jobList para -ram=16g create jobList para try ... check ... push ... etc... # Completed: 99 of 99 jobs # CPU time in finished jobs: 52719s 878.64m 14.64h 0.61d 0.002 y # IO & Wait Time: 1804s 30.07m 0.50h 0.02d 0.000 y # Average job time: 551s 9.18m 0.15h 0.01d # Longest finished job: 1397s 23.28m 0.39h 0.02d # Submission to last job: 1487s 24.78m 0.41h 0.02d # construct a list of all possible maf file names. # they do not all exist in each of the species directories find . -type f | grep "maf.gz" | wc -l # 59599 find . -type f | grep ".maf.gz$" | xargs -L 1 basename | sort -u \ > run.maf.list wc -l run.maf.list # 678 run.maf.list # number of chroms with data: awk -F'.' '{print $1}' run.maf.list | sed -e 's/hg38_//;' \ | sort | uniq -c | sort -n | wc -l # 358 mkdir /hive/data/genomes/hg38/bed/multiz100way/splitRun cd /hive/data/genomes/hg38/bed/multiz100way/splitRun mkdir maf run cd run mkdir penn cp -p /cluster/bin/penn/multiz.2009-01-21_patched/multiz penn cp -p /cluster/bin/penn/multiz.2009-01-21_patched/maf_project penn cp -p /cluster/bin/penn/multiz.2009-01-21_patched/autoMZ penn # set the db and pairs directories here cat > autoMultiz.csh << '_EOF_' #!/bin/csh -ef set db = hg38 set c = $1 set result = $2 set run = `/bin/pwd` set tmp = /dev/shm/$db/multiz.$c set pairs = /hive/data/genomes/hg38/bed/multiz100way/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 $tmp /bin/rmdir --ignore-fail-on-non-empty /dev/shm/$db '_EOF_' # << happy emacs chmod +x autoMultiz.csh cat << '_EOF_' > template #LOOP ./autoMultiz.csh $(file1) {check out line+ /hive/data/genomes/hg38/bed/multiz100way/splitRun/maf/$(root1).maf} #ENDLOOP '_EOF_' # << happy emacs sed -e 's/.gz//;' ../../mafSplit/run.maf.list > maf.list ssh ku cd /hive/data/genomes/hg38/bed/multiz100way/splitRun/run gensub2 maf.list single template jobList para create jobList # Completed: 678 of 678 jobs # CPU time in finished jobs: 10288136s 171468.93m 2857.82h 119.08d 0.326 y # IO & Wait Time: 17393s 289.89m 4.83h 0.20d 0.001 y # Average job time: 15200s 253.33m 4.22h 0.18d # Longest finished job: 103429s 1723.82m 28.73h 1.20d # Submission to last job: 297652s 4960.87m 82.68h 3.45d # put the split maf results back together into a single per-chrom maf file # eliminate duplicate comments ssh hgwdev cd /hive/data/genomes/hg38/bed/multiz100way/splitRun mkdir ../maf # no need to save the comments since they are lost with mafAddIRows cat << '_EOF_' >> runOne #!/bin/csh -fe set C = $1 if ( -s ../maf/${C}.maf.gz ) then rm -f ../maf/${C}.maf.gz endif if ( -s maf/hg38_${C}.00.maf ) then head -q -n 1 maf/hg38_${C}.00.maf | sort -u > ../maf/${C}.maf grep -h -v "^#" `ls maf/hg38_${C}.*.maf | sort -t. -k2,2n` >> ../maf/${C}.maf tail -q -n 1 maf/hg38_${C}.00.maf | sort -u >> ../maf/${C}.maf else touch ../maf/${C}.maf endif '_EOF_' # << happy emacs chmod +x runOne cat << '_EOF_' >> template #LOOP runOne $(root1) {check out exists ../maf/$(root1).maf} #ENDLOOP '_EOF_' # << happy emacs cut -f1 ../../../chrom.sizes > chr.list ssh ku cd /hive/data/genomes/hg38/bed/multiz100way/splitRun gensub2 chr.list single template jobList para -ram=16g create jobList para try ... check ... push ... etc ... para -maxJob=32 push # Completed: 455 of 455 jobs # CPU time in finished jobs: 1040s 17.33m 0.29h 0.01d 0.000 y # IO & Wait Time: 4118s 68.63m 1.14h 0.05d 0.000 y # Average job time: 11s 0.19m 0.00h 0.00d # Longest finished job: 238s 3.97m 0.07h 0.00d # Submission to last job: 241s 4.02m 0.07h 0.00d # 97 of them have empty results, they have to be removed ls -ogrt | awk '$3 == 0' | awk '{print $NF}' | xargs rm -f # Load into database mkdir -p /gbdb/hg38/multiz100way/maf cd /hive/data/genomes/hg38/bed/multiz100way/maf ln -s `pwd`/*.maf /gbdb/hg38/multiz100way/maf/ # this generates an immense multiz100way.tab file in the directory # where it is running. Best to run this over in scratch. # This is going to take all day. cd /dev/shm time hgLoadMaf -pathPrefix=/gbdb/hg38/multiz100way/maf hg38 multiz100way # Loaded 114640349 mafs in 358 files from /gbdb/hg38/multiz100way/maf # real 101m18.944s # -rw-rw-r-- 1 6198828190 May 5 12:33 multiz100way.tab wc -l multiz100way.tab # 114640349 multiz100way.tab time (cat /gbdb/hg38/multiz100way/maf/*.maf \ | hgLoadMafSummary -verbose=2 -minSize=30000 \ -mergeGap=1500 -maxSize=200000 hg38 multiz100waySummary stdin) # Created 18099188 summary blocks from 4120456307 components and 114640349 mafs from stdin # real 173m32.469s wc -l multiz100way*.tab # 114640349 multiz100way.tab # 18099188 multiz100waySummary.tab rm multiz100way*.tab ####################################################################### # GAP ANNOTATE MULTIZ9WAY MAF AND LOAD TABLES (DONE - 2015-05-06 - Hiram) # mafAddIRows has to be run on single chromosome maf files, it does not # function correctly when more than one reference sequence # are in a single file. mkdir -p /hive/data/genomes/hg38/bed/multiz100way/anno cd /hive/data/genomes/hg38/bed/multiz100way/anno # check for N.bed files everywhere: for DB in `cat ../species.list` do if [ ! -s /hive/data/genomes/${DB}/${DB}.N.bed ]; then echo "MISS: ${DB}" cd /hive/data/genomes/${DB} twoBitInfo -nBed ${DB}.2bit ${DB}.N.bed else echo " OK: ${DB}" fi done cd /hive/data/genomes/hg38/bed/multiz100way/anno for DB in `cat ../species.list` do echo "${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 *.bed | wc -l # 100 screen -S hg38 # use a screen to control this longish job ssh ku cd /hive/data/genomes/hg38/bed/multiz100way/anno mkdir result cat << '_EOF_' > template #LOOP mafAddIRows -nBeds=nBeds $(path1) /hive/data/genomes/hg38/hg38.2bit {check out line+ result/$(file1)} #ENDLOOP '_EOF_' # << happy emacs ls ../maf/*.maf > maf.list gensub2 maf.list single template jobList # no need to limit these jobs, there are only 93 of them para -ram=64g create jobList para try ... check ... para -maxJob=10 push # Completed: 348 of 358 jobs # Crashed: 10 jobs # CPU time in finished jobs: 21567s 359.46m 5.99h 0.25d 0.001 y # IO & Wait Time: 1252s 20.86m 0.35h 0.01d 0.000 y # Average job time: 66s 1.09m 0.02h 0.00d # Longest finished job: 2593s 43.22m 0.72h 0.03d # Submission to last job: 5232s 87.20m 1.45h 0.06d # running the last 10 jobs on kolossus with 200 Gb memory limit: ssh kolossus cd /hive/data/genomes/hg38/bed/multiz100way/anno cat << '_EOF_' > bigJobs.sh #!/bin/sh # 200 Gb memory limits export sizeG=200000000 ulimit -d $sizeG ulimit -v $sizeG mafAddIRows -nBeds=nBeds ../maf/chr1.maf /hive/data/genomes/hg38/hg38.2bit result/chr1.maf & mafAddIRows -nBeds=nBeds ../maf/chr2.maf /hive/data/genomes/hg38/hg38.2bit result/chr2.maf & mafAddIRows -nBeds=nBeds ../maf/chr3.maf /hive/data/genomes/hg38/hg38.2bit result/chr3.maf wait mafAddIRows -nBeds=nBeds ../maf/chr4.maf /hive/data/genomes/hg38/hg38.2bit result/chr4.maf & mafAddIRows -nBeds=nBeds ../maf/chr5.maf /hive/data/genomes/hg38/hg38.2bit result/chr5.maf & mafAddIRows -nBeds=nBeds ../maf/chr6.maf /hive/data/genomes/hg38/hg38.2bit result/chr6.maf wait mafAddIRows -nBeds=nBeds ../maf/chr7.maf /hive/data/genomes/hg38/hg38.2bit result/chr7.maf & mafAddIRows -nBeds=nBeds ../maf/chr8.maf /hive/data/genomes/hg38/hg38.2bit result/chr8.maf & mafAddIRows -nBeds=nBeds ../maf/chr9.maf /hive/data/genomes/hg38/hg38.2bit result/chr9.maf wait mafAddIRows -nBeds=nBeds ../maf/chrX.maf /hive/data/genomes/hg38/hg38.2bit result/chrX.maf '_EOF_' # << happy emacs chmod +x bigJobs.sh time (./bigJobs.sh) > bigJobs.log 2>&1 # real 203m2.858s du -hsc result # 789G result # Load into database rm -f /gbdb/hg38/multiz100way/maf/* cd /hive/data/genomes/hg38/bed/multiz100way/anno/result ln -s `pwd`/*.maf /gbdb/hg38/multiz100way/maf/ # this generates an immense multiz100way.tab file in the directory # where it is running. Best to run this over in scratch. # This is going to take all day. cd /dev/shm time hgLoadMaf -pathPrefix=/gbdb/hg38/multiz100way/maf hg38 multiz100way # Loaded 114670205 mafs in 358 files from /gbdb/hg38/multiz100way/maf # real 193m17.801s # -rw-rw-r-- 1 6239176446 May 6 19:14 multiz100way.tab time (cat /gbdb/hg38/multiz100way/maf/*.maf \ | hgLoadMafSummary -verbose=2 -minSize=30000 \ -mergeGap=1500 -maxSize=200000 hg38 multiz100waySummary stdin) # Created 18099188 summary blocks from 4120456307 components # and 114670205 mafs from stdin # real 270m53.048s # This process became very large: # PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND # 4328 hiram 20 0 165g 165g 1764 R 96.1 16.4 249:34.23 hgLoadMafSummary # even though the .tab file isn't gigantic: # -rw-rw-r-- 1 894587030 May 7 13:51 multiz100waySummary.tab wc -l multiz100way*.tab # 114670205 multiz100way.tab # 18099188 multiz100waySummary.tab rm multiz100way*.tab ############################################################################## # MULTIZ7WAY MAF FRAMES (DONE - 2015-05-05 - Hiram) ssh hgwdev mkdir /hive/data/genomes/hg38/bed/multiz100way/frames cd /hive/data/genomes/hg38/bed/multiz100way/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 hg38 -N -e \ "select id from organism where name='$orgName'"` if ($orgId == "") then echo "Mrnas: 0" else set count = `hgsql hg38 -N -e "select count(*) from gbCdnaInfo where organism=$orgId"` echo "Mrnas: ${count}" endif end '_EOF_' # << happy emacs chmod +x ./showGenes.csh time ./showGenes.csh # rearrange that output to create four sections, and place these names # in .list files here: # 1. knownGene: hg38 mm10 # 2. refGene: bosTau8 danRer10 macFas5 rheMac3 rn6 xenTro7 # 3. ensGene: ailMel1 anaPla1 anoCar2 astMex1 calJac3 canFam3 cavPor3 # dasNov3 equCab2 fr3 gadMor1 galGal4 gasAcu1 gorGor3 latCha1 lepOcu1 # loxAfr3 melGal1 monDom5 musFur1 myoLuc2 ornAna1 oryCun2 oryLat2 # otoGar3 oviAri3 panTro4 papAnu2 pelSin1 petMar2 ponAbe2 pteVam1 # sarHar1 speTri2 susScr3 taeGut2 tetNig2 xipMac1 # 4. no annotation: criGri1 chiLan1 vicPac2 felCat8 allMis1 oreNil2 # nomLeu3 chlSab2 saiBol1 hetGla2 ochPri3 turTru2 cerSim1 eriEur2 # sorAra2 triMan1 echTel2 macEug2 geoFor1 melUnd1 chrPic2 # tupChi1 odoRosDiv1 jacJac1 micOch1 mesAur1 octDeg1 camFer1 # orcOrc1 panHod1 capHir1 lepWed1 pteAle1 eptFus1 myoDav1 # conCri1 eleEdw1 chrAsi1 oryAfe1 colLiv1 falChe1 falPer1 # ficAlb2 zonAlb1 pseHum1 amaVit1 araMac1 cheMyd1 apaSpi1 # takFla1 neoBri1 hapBur1 mayZeb1 punNye1 mkdir genes # 1. knownGene: hg38 and mm10 for DB in hg38 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 echo -n "# ${DB}: " genePredCheck -db=${DB} genes/${DB}.gp.gz done # hg38: checked: 21179 failed: 0 # mm10: checked: 21036 failed: 0 # 2. refGene: bosTau8 danRer10 macFas5 rheMac3 rn6 xenTro7 # want the full extended genePred: for DB in `cat refGene.list` do hgsql -N -e "select * from refGene" ${DB} | cut -f2- \ | genePredSingleCover stdin stdout | gzip -2c \ > /scratch/tmp/${DB}.tmp.gz mv /scratch/tmp/${DB}.tmp.gz genes/$DB.gp.gz echo -n "# ${DB}: " genePredCheck -db=${DB} genes/${DB}.gp.gz done # bosTau8: checked: 13329 failed: 0 # danRer10: checked: 14201 failed: 0 # macFas5: checked: 2196 failed: 0 # rheMac3: checked: 5804 failed: 0 # rn6: checked: 17140 failed: 0 # xenTro7: checked: 8502 failed: 0 # 3. ensGene, want the full extended genePred: for DB in `cat ensGene.list` do hgsql -N -e "select * from ensGene" ${DB} | cut -f2- \ | genePredSingleCover stdin stdout | gzip -2c \ > /scratch/tmp/${DB}.tmp.gz mv /scratch/tmp/${DB}.tmp.gz genes/$DB.gp.gz echo -n "# ${DB}: " genePredCheck -db=${DB} genes/${DB}.gp.gz done # ailMel1: checked: 19204 failed: 0 # anaPla1: checked: 15482 failed: 0 # anoCar2: checked: 18532 failed: 0 # astMex1: checked: 22979 failed: 0 # calJac3: checked: 20827 failed: 0 # canFam3: checked: 19507 failed: 0 # cavPor3: checked: 18631 failed: 0 # dasNov3: checked: 22586 failed: 0 # equCab2: checked: 20412 failed: 0 # fr3: checked: 18014 failed: 0 # gadMor1: checked: 76607 failed: 0 # galGal4: checked: 15391 failed: 0 # gasAcu1: checked: 20631 failed: 0 # gorGor3: checked: 20758 failed: 0 # latCha1: checked: 19539 failed: 0 # lepOcu1: checked: 18252 failed: 0 # loxAfr3: checked: 19986 failed: 0 # melGal1: checked: 14050 failed: 0 # monDom5: checked: 21033 failed: 0 # musFur1: checked: 19626 failed: 0 # myoLuc2: checked: 19685 failed: 0 # ornAna1: checked: 21311 failed: 0 # oryCun2: checked: 19165 failed: 0 # oryLat2: checked: 19586 failed: 0 # otoGar3: checked: 19472 failed: 0 # oviAri3: checked: 20793 failed: 0 # panTro4: checked: 18657 failed: 0 # papAnu2: checked: 18904 failed: 0 # pelSin1: checked: 18093 failed: 0 # petMar2: checked: 10381 failed: 0 # ponAbe2: checked: 20220 failed: 0 # pteVam1: checked: 23293 failed: 0 # sarHar1: checked: 18663 failed: 0 # speTri2: checked: 18796 failed: 0 # susScr3: checked: 21596 failed: 0 # taeGut2: checked: 17247 failed: 0 # tetNig2: checked: 19539 failed: 0 # xipMac1: checked: 20320 failed: 0 # kluster job to annotate each maf file screen -S hg38 # manage long running procedure with screen ssh ku cd /hive/data/genomes/hg38/bed/multiz100way/frames cat << '_EOF_' > runOne #!/bin/csh -fe set C = $1 set G = $2 cat ../maf/${C}.maf | genePredToMafFrames hg38 stdin stdout \ ${G} genes/${G}.gp.gz | gzip > parts/${C}.${G}.mafFrames.gz '_EOF_' # << happy emacs chmod +x runOne ls ../maf | sed -e "s/.maf//" > chr.list ls genes | sed -e "s/.gp.gz//" > gene.list cat << '_EOF_' > template #LOOP runOne $(root1) $(root2) {check out exists+ parts/$(root1).$(root2).mafFrames.gz} #ENDLOOP '_EOF_' # << happy emacs mkdir parts gensub2 chr.list gene.list template jobList para -ram=64g create jobList para try ... check ... push # Completed: 16468 of 16468 jobs # CPU time in finished jobs: 202898s 3381.64m 56.36h 2.35d 0.006 y # IO & Wait Time: 129253s 2154.21m 35.90h 1.50d 0.004 y # Average job time: 20s 0.34m 0.01h 0.00d # Longest finished job: 3278s 54.63m 0.91h 0.04d # Submission to last job: 4102s 68.37m 1.14h 0.05d # collect all results into one file: cd /hive/data/genomes/hg38/bed/multiz100way/frames time find ./parts -type f | while read F do echo "${F}" 1>&2 zcat ${F} done | sort -k1,1 -k2,2n > multiz100wayFrames.bed # real 4m35.000s # -rw-rw-r-- 1 1175092685 May 7 10:50 multiz100wayFrames.bed gzip multiz100wayFrames.bed # verify there are frames on everything, should be 46 species: # (count from: ls genes | wc) zcat multiz100wayFrames.bed.gz | awk '{print $4}' | sort | uniq -c \ | sed -e 's/^/# /;' > species.check.list wc -l species.check.list # 46 # 258752 ailMel1 # 503273 anaPla1 # 567017 anoCar2 # 611070 astMex1 # 164101 bosTau8 # 246294 calJac3 # 286818 canFam3 # 245975 cavPor3 # 381305 danRer10 # 281610 dasNov3 # 255526 equCab2 # 556448 fr3 # 471583 gadMor1 # 681358 galGal4 # 623263 gasAcu1 # 196510 gorGor3 # 207288 hg38 # 437877 latCha1 # 736191 lepOcu1 # 257755 loxAfr3 # 16693 macFas5 # 562685 melGal1 # 258209 mm10 # 676384 monDom5 # 278050 musFur1 # 251293 myoLuc2 # 663051 ornAna1 # 250416 oryCun2 # 546870 oryLat2 # 244848 otoGar3 # 298167 oviAri3 # 200545 panTro4 # 190478 papAnu2 # 551703 pelSin1 # 316423 petMar2 # 222799 ponAbe2 # 237710 pteVam1 # 50409 rheMac3 # 199469 rn6 # 569235 sarHar1 # 232093 speTri2 # 206517 susScr3 # 623969 taeGut2 # 641809 tetNig2 # 231941 xenTro7 # 562318 xipMac1 # load the resulting file ssh hgwdev cd /hive/data/genomes/hg38/bed/multiz100way/frames time hgLoadMafFrames hg38 multiz100wayFrames multiz100wayFrames.bed.gz # real 3m14.812s hgsql -e 'select count(*) from multiz100wayFrames;' hg38 # +----------+ # | count(*) | # +----------+ # | 17054098 | # +----------+ time featureBits -countGaps hg38 multiz100wayFrames # 82169754 bases of 3209286105 (2.560%) in intersection # real 1m56.332s # enable the trackDb entries: # frames multiz100wayFrames # irows on # appears to work OK ######################################################################### # Phylogenetic tree from 100-way (DONE - 2013-09-13 - Hiram) mkdir /hive/data/genomes/hg38/bed/multiz100way/4d cd /hive/data/genomes/hg38/bed/multiz100way/4d # the annotated maf's are in: ../anno/result/*.maf # using knownGene for hg38, only transcribed genes and nothing # from the randoms and other misc. hgsql -Ne "select name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds from knownGene where cdsEnd > cdsStart;" hg38 \ | egrep -E -v "chrM|chrUn|random|_alt" > knownGene.gp wc -l *.gp # 93567 knownGene.gp # verify it is only on the chroms: cut -f2 knownGene.gp | sort | uniq -c | sort -rn | sed -e 's/^/ # /;' # 7789 chr1 # 7329 chr19 # 6461 chr17 # 6289 chr11 # 6136 chr2 # 5760 chr12 # 5585 chr3 # 4930 chr16 # 4277 chr7 # 4214 chr6 # 4024 chr5 # 3756 chr14 # 3546 chr4 # 3513 chr8 # 3286 chr15 # 2896 chrX # 2864 chr9 # 2847 chr10 # 2102 chr22 # 2018 chr20 # 1651 chr18 # 1135 chr13 # 930 chr21 # 229 chrY genePredSingleCover knownGene.gp stdout | sort > knownGeneNR.gp wc -l knownGeneNR.gp # 19175 knownGeneNR.gp ssh ku mkdir /hive/data/genomes/hg38/bed/multiz100way/4d/run cd /hive/data/genomes/hg38/bed/multiz100way/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/hg38/bed/multiz100way" set c = $1 set infile = $r/anno/result/$2 set outfile = $3 cd /dev/shm # 'clean' maf, removes all chrom names, leaves only the db name perl -wpe 's/^s ([^.]+)\.\S+/s $1/' $infile > $c.maf awk -v C=$c '$2 == C {print}' $r/4d/knownGeneNR.gp | sed -e "s/\t$c\t/\thg38.$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 $c.maf -o SS > $c.ss $PHASTBIN/msa_view -i SS --tuple-size 1 $c.ss > $r/4d/run/$outfile else echo "" > $r/4d/run/$outfile endif rm -f $c.gp $c.maf $c.ss '_EOF_' # << happy emacs chmod +x 4d.csh ls -1S /hive/data/genomes/hg38/bed/multiz100way/anno/result/*.maf \ | sed -e "s#.*multiz100way/anno/result/##" \ | egrep -E -v "chrM|chrUn|random|_alt" > maf.list cat << '_EOF_' > template #LOOP 4d.csh $(root1) $(path1) {check out line+ ../mfa/$(root1).mfa} #ENDLOOP '_EOF_' # << happy emacs gensub2 maf.list single template jobList para -ram=64g create jobList para try ... check ... push ... etc... para time # Completed: 24 of 24 jobs # CPU time in finished jobs: 34007s 566.79m 9.45h 0.39d 0.001 y # IO & Wait Time: 6873s 114.54m 1.91h 0.08d 0.000 y # Average job time: 1703s 28.39m 0.47h 0.02d # Longest finished job: 3815s 63.58m 1.06h 0.04d # Submission to last job: 3830s 63.83m 1.06h 0.04d # combine mfa files ssh hgwdev cd /hive/data/genomes/hg38/bed/multiz100way/4d # verify no tiny files: ls -og mfa | sort -k3nr | tail -1 # -rw-rw-r-- 1 283090 May 7 09:42 chrY.mfa #want comma-less species.list time /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/msa_view \ --aggregate "`cat ../species.list`" mfa/*.mfa | sed s/"> "/">"/ \ > 4d.all.mfa # real 0m4.406s # check they are all in there: grep "^>" 4d.all.mfa | wc -l # 100 sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \ hg38.100way.nh sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \ ../hg38.100way.nh > tree-commas.nh # use phyloFit to create tree model (output is phyloFit.mod) time /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 424m51.691s mv phyloFit.mod all.mod grep TREE all.mod # compare these calculated lengths to what the hg19 process made: grep TREE /hive/data/genomes/hg19/bed/multiz100way/4d/all.mod \ | sed -e 's/TREE: //; s/bosTau7/bosTau8/; s/chlSab1/chlSab2/; s/chrPic1/chrPic2/; s/danRer7/danRer10/; s/felCat5/felCat8/; s/papHam1/papAnu2/; s/rn5/rn6/; s/hg19/hg38/;' \ | /cluster/bin/phast/all_dists /dev/stdin | grep hg38 \ | sed -e "s/hg38.//;" | sort > hg19.dists grep TREE all.mod | sed -e 's/TREE: //;' \ | /cluster/bin/phast/all_dists /dev/stdin | grep hg38 \ | sed -e "s/hg38.//;" | sort > hg38.dists # printing out the 'hg19', the 'hg38' the 'difference' and # percent difference/delta join hg19.dists hg38.dists | awk '{ printf "#\t%s\t%8.6f\t%8.6f\t%8.6f\t%8.6f\n", $1, $2, $3, $2-$3, 100*($2-$3)/$3 }' \ | sort -k6n | head -6 # top 6 percent changes: # db hg19 hg38 hg19-hg38 percent delta # macEug2 0.778331 0.838209 -0.059878 -7.143564 # ornAna1 0.967705 1.007930 -0.040225 -3.990853 # araMac1 1.275902 1.309637 -0.033735 -2.575905 # amaVit1 1.259843 1.286088 -0.026245 -2.040685 # zonAlb1 1.375094 1.390649 -0.015555 -1.118542 # melUnd1 1.234923 1.246845 -0.011922 -0.956173 # and the seven new assemblies: # | egrep "bosTau|chlSab|chrPic|danRer|felCat|papAnu|rn6" # db hg19 hg38 hg19-hg38 percent delta # bosTau8 0.433906 0.433893 0.000013 0.002996 # chlSab2 0.070327 0.070253 0.000074 0.105334 # chrPic2 1.004158 0.996918 0.007240 0.726238 # danRer10 2.211308 2.209038 0.002270 0.102760 # felCat8 0.347688 0.346784 0.000904 0.260681 # papAnu2 0.069896 0.069599 0.000297 0.426730 # rn6 0.520202 0.520167 0.000035 0.006729 ######################################################################### # phastCons 100-way (DONE - 2015-05-07 - Hiram) # split 100way mafs into 10M chunks and generate sufficient statistics # files for # phastCons ssh ku mkdir -p /hive/data/genomes/hg38/bed/multiz100way/cons/ss mkdir -p /hive/data/genomes/hg38/bed/multiz100way/cons/msa.split cd /hive/data/genomes/hg38/bed/multiz100way/cons/msa.split cat << '_EOF_' > doSplit.csh #!/bin/csh -ef set c = $1 set MAF = /hive/data/genomes/hg38/bed/multiz100way/anno/result/$c.maf set WINDOWS = /hive/data/genomes/hg38/bed/multiz100way/cons/ss/$c set WC = `cat $MAF | wc -l` set NL = `grep "^#" $MAF | wc -l` if ( -s $2 ) then exit 0 endif if ( -s $2.running ) then exit 0 endif date >> $2.running rm -fr $WINDOWS mkdir $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 date >> $2 rm -f $2.running '_EOF_' # << happy emacs chmod +x doSplit.csh cat << '_EOF_' > template #LOOP doSplit.csh $(root1) {check out line+ $(root1).done} #ENDLOOP '_EOF_' # << happy emacs # do the easy ones first to see some immediate results ls -1S -r ../../anno/result | sed -e "s/.maf//;" > maf.list # all can finish OK at a 64Gb memory limit gensub2 maf.list single template jobList para -ram=64g create jobList para try ... check ... etc para push # Completed: 358 of 358 jobs # CPU time in finished jobs: 125184s 2086.40m 34.77h 1.45d 0.004 y # IO & Wait Time: 6665s 111.08m 1.85h 0.08d 0.000 y # Average job time: 368s 6.14m 0.10h 0.00d # Longest finished job: 19073s 317.88m 5.30h 0.22d # Submission to last job: 19138s 318.97m 5.32h 0.22d # 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 ku mkdir -p /hive/data/genomes/hg38/bed/multiz100way/cons/run.cons cd /hive/data/genomes/hg38/bed/multiz100way/cons/run.cons # This is setup for multiple runs based on subsets, but only running # the 'all' subset here. # 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 set PHASTBIN = /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin set c = $1 set f = $2 set len = $3 set cov = $4 set rho = $5 set grp = $cwd:t set cons = /hive/data/genomes/hg19/bed/multiz100way/cons set tmp = $cons/tmp/$f mkdir -p $tmp set ssSrc = $cons/ss set useGrp = "$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 $ssSrc/$c/$f.ss $tmp else ln -s $ssSrc/$c/$f.ss $tmp ln -s $cons/$grp/$grp.mod $tmp endif pushd $tmp > /dev/null if (-s $grp.non-inf) then $PHASTBIN/phastCons $f.ss $useGrp \ --rho $rho --expected-length $len --target-coverage $cov --quiet \ --not-informative `cat $grp.non-inf` \ --seqname $c --idpref $c --most-conserved $f.bed --score > $f.pp else $PHASTBIN/phastCons $f.ss $useGrp \ --rho $rho --expected-length $len --target-coverage $cov --quiet \ --seqname $c --idpref $c --most-conserved $f.bed --score > $f.pp endif popd > /dev/null mkdir -p pp/$c bed/$c sleep 4 touch pp/$c bed/$c rm -f pp/$c/$f.pp rm -f bed/$c/$f.bed mv $tmp/$f.pp pp/$c mv $tmp/$f.bed bed/$c rm -fr $tmp '_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) $(file1) 45 0.3 0.3 {check out line+ pp/$(root1)/$(file1).pp} #ENDLOOP '_EOF_' # << happy emacs ls -1S ../ss/chr*/chr* | sed -e "s/.ss$//" > ss.list wc -l ss.list # 644 ss.list # Create parasol batch and run it # run for all species cd /hive/data/genomes/hg38/bed/multiz100way/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 # beware overwhelming the cluster with these fast running high I/O jobs para -ram=32g create jobList para try ... check ... para -maxJob=16 push # Completed: 644 of 644 jobs # CPU time in finished jobs: 100849s 1680.82m 28.01h 1.17d 0.003 y # IO & Wait Time: 4332s 72.19m 1.20h 0.05d 0.000 y # Average job time: 163s 2.72m 0.05h 0.00d # Longest finished job: 399s 6.65m 0.11h 0.00d # Submission to last job: 4087s 68.12m 1.14h 0.05d # create Most Conserved track cd /hive/data/genomes/hg38/bed/multiz100way/cons/all time cut -f1 ../../../../chrom.sizes | while read C do echo $C 1>&2 ls -d bed/${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 1m23.557s # -rw-rw-r-- 1 357720691 May 8 08:14 tmpMostConserved.bed time /cluster/bin/scripts/lodToBedScore tmpMostConserved.bed \ > mostConserved.bed # real 1m33.908s # -rw-rw-r-- 1 366712427 May 8 08:24 mostConserved.bed # load into database ssh hgwdev cd /hive/data/genomes/hg38/bed/multiz100way/cons/all time hgLoadBed hg38 phastConsElements100way mostConserved.bed # Read 10350729 elements of size 5 from mostConserved.bed # real 1m56.076s # --rho 0.3 --expected-length 45 --target-coverage 0.3 time featureBits hg38 -enrichment knownGene:cds phastConsElements100way # knownGene:cds 1.260%, phastConsElements100way 5.319%, both 0.826%, cover 65.55%, enrich 12.33x # real 1m12.540s # Try for 5% overall cov, and 70% CDS cov time featureBits hg38 -enrichment refGene:cds phastConsElements100way # refGene:cds 1.210%, phastConsElements100way 5.319%, both 0.817%, cover 67.50%, enrich 12.69x # real 1m10.541s # Create merged posterier probability file and wiggle track data files cd /hive/data/genomes/hg38/bed/multiz100way/cons/all mkdir downloads time for D in `ls -d pp/chr* | sed -e 's#pp/##'` do echo "working: $D" 1>&2 find ./pp/${D} -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/${D}.phastCons100way.wigFix.gz done # real 25m35.972s # encode those files into wiggle data time (zcat downloads/*.wigFix.gz \ | wigEncode stdin phastCons100way.wig phastCons100way.wib) # Converted stdin, upper limit 1.00, lower limit 0.00 # real 13m36.113s du -hsc *.wi? # 2.8G phastCons100way.wib # 283M phastCons100way.wig # encode into a bigWig file: # (warning wigToBigWig process may be too large for memory limits # in bash, to avoid the 32 Gb memory limit, set 180 Gb here: sizeG=188743680 export sizeG ulimit -d $sizeG ulimit -v $sizeG time (zcat downloads/*.wigFix.gz \ | wigToBigWig stdin ../../../../chrom.sizes phastCons100way.bw) # real 39m3.769s # -rw-rw-r-- 1 5886377734 May 8 09:57 phastCons100way.bw bigWigInfo phastCons100way.bw version: 4 isCompressed: yes isSwapped: 0 primaryDataSize: 4,001,253,148 primaryIndexSize: 93,019,356 zoomLevels: 10 chromCount: 356 basesCovered: 2,944,992,024 mean: 0.098095 min: 0.000000 max: 1.000000 std: 0.233810 # if you wanted to use the bigWig file, loading bigWig table: # but we don't use the bigWig file mkdir /gbdb/hg38/bbi ln -s `pwd`/phastCons100way.bw /gbdb/hg38/bbi hgsql hg38 -e 'drop table if exists phastCons100way; \ create table phastCons100way (fileName varchar(255) not null); \ insert into phastCons100way values ("/gbdb/hg38/bbi/phastCons100way.bw");' # Load gbdb and database with wiggle. ssh hgwdev cd /hive/data/genomes/hg38/bed/multiz100way/cons/all ln -s `pwd`/phastCons100way.wib /gbdb/hg38/multiz100way/phastCons100way.wib time hgLoadWiggle -pathPrefix=/gbdb/hg38/multiz100way hg38 \ phastCons100way phastCons100way.wig # real 0m31.883s time wigTableStats.sh hg38 phastCons100way # real 0m13.628s # db.table min max mean count sumData hg38.phastCons100way 0 1 0.0980951 2944992024 2.88889e+08 # stdDev viewLimits # 0.23381 viewLimits=0:1 # Create histogram to get an overview of all the data ssh hgwdev cd /hive/data/genomes/hg38/bed/multiz100way/cons/all time hgWiggle -doHistogram -db=hg38 \ -hBinSize=0.001 -hBinCount=1000 -hMinVal=0.0 -verbose=2 \ phastCons100way > histogram.data 2>&1 # real 2m38.952s # 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 " Human Hg38 Histogram phastCons100way track" set xlabel " phastCons100way 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 # complains about font, but makes the png image: # Could not find/open font when opening font "arial", using internal non-scalable font display histo.png & ######################################################################### # phyloP for 100-way (DONE - 2015-05-08 - Hiram) # # all vertebrates # # split SS files into 1M chunks, this business needs smaller files # to complete ssh ku mkdir /hive/data/genomes/hg38/bed/multiz100way/consPhyloP cd /hive/data/genomes/hg38/bed/multiz100way/consPhyloP mkdir ss run.split cd run.split cat << '_EOF_' > doSplit.csh #!/bin/csh -ef set c = $1 set MAF = /hive/data/genomes/hg38/bed/multiz100way/anno/result/$c.maf set WINDOWS = /hive/data/genomes/hg38/bed/multiz100way/consPhyloP/ss/$c set WC = `cat $MAF | wc -l` set NL = `grep "^#" $MAF | wc -l` if ( -s $2 ) then exit 0 endif if ( -s $2.running ) then exit 0 endif date >> $2.running rm -fr $WINDOWS 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 1000000,0 -I 1000 -B 5000 endif popd > /dev/null date >> $2 rm -f $2.running '_EOF_' # << happy emacs chmod +x doSplit.csh # do the easy ones first to see some immediate results ls -1S -r ../../anno/result | sed -e "s/.maf//;" > maf.list # this needs a {check out line+ $(root1.done)} test for verification: cat << '_EOF_' > template #LOOP ./doSplit.csh $(root1) $(root1).done #ENDLOOP '_EOF_' # << happy emacs gensub2 maf.list single template jobList # all can complete successfully at the 64Gb memory limit para -ram=64g create jobList para try ... check ... push ... etc... # largest one (chr2) becomes this large near its end: # PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND # 31927 hiram 20 0 55.3g 54g 788 R 99.5 21.7 317:45.12 msa_split # Completed: 358 of 358 jobs # CPU time in finished jobs: 125744s 2095.74m 34.93h 1.46d 0.004 y # IO & Wait Time: 6413s 106.88m 1.78h 0.07d 0.000 y # Average job time: 369s 6.15m 0.10h 0.00d # Longest finished job: 20020s 333.67m 5.56h 0.23d # Submission to last job: 20147s 335.78m 5.60h 0.23d # run phyloP with score=LRT ssh ku mkdir /cluster/data/hg38/bed/multiz100way/consPhyloP cd /cluster/data/hg38/bed/multiz100way/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 ../../4d/all.mod | awk '{printf "%0.3f\n", $3 + $4}' # 0.508 /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/modFreqs \ ../../4d/all.mod 0.508 > all.mod # verify, the BACKGROUND should now be paired up: grep BACK all.mod # BACKGROUND: 0.246000 0.254000 0.254000 0.246000 cat << '_EOF_' > doPhyloP.csh #!/bin/csh -fe set PHASTBIN = /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin set f = $1 set ssFile = $1:t set out = $2 set cName = $f:h set n = $f:r:e set grp = $cwd:t set cons = /hive/data/genomes/hg38/bed/multiz100way/consPhyloP set tmp = $cons/tmp/$grp/$f /bin/rm -fr $tmp /bin/mkdir -p $tmp set ssSrc = "$cons/ss/$cName/$ssFile" set useGrp = "$grp.mod" /bin/ln -s $cons/run.phyloP/$grp.mod $tmp pushd $tmp > /dev/null echo source: $ssSrc.ss $PHASTBIN/phyloP --method LRT --mode CONACC --wig-scores --chrom $cName \ -i SS $useGrp $ssSrc.ss > $ssFile.wigFix popd > /dev/null /bin/mkdir -p $out:h sleep 4 /bin/touch $out:h /bin/mv $tmp/$ssFile.wigFix $out /bin/rm -fr $tmp /bin/rmdir --ignore-fail-on-non-empty $cons/tmp/$grp /bin/rmdir --ignore-fail-on-non-empty $cons/tmp '_EOF_' # << happy emacs chmod +x doPhyloP.csh # Create list of chunks find ../ss -type f | sed -e "s/.ss$//; s#../ss/##;" > ss.list # make sure the list looks good wc -l ss.list # 645 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/hg38/bed/multiz100way/consPhyloP/all cd /hive/data/genomes/hg38/bed/multiz100way/consPhyloP/all rm -fr wigFix mkdir wigFix gensub2 ../run.phyloP/ss.list single ../run.phyloP/template jobList # beware overloading the cluster with these quick and high I/O jobs para -ram=32g create jobList para try ... check ... para -maxJob=16 push para time > run.time # Completed: 3310 of 3310 jobs # CPU time in finished jobs: 4871081s 81184.68m 1353.08h 56.38d 0.154 y # IO & Wait Time: 24135s 402.25m 6.70h 0.28d 0.001 y # Average job time: 1479s 24.65m 0.41h 0.02d # Longest finished job: 2691s 44.85m 0.75h 0.03d # Submission to last job: 50957s 849.28m 14.15h 0.59d mkdir downloads time for D in `ls -d wigFix/chr* | sed -e 's#wigFix/##'` do echo "working: $D" 1>&2 find ./wigFix/${D} -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/${D}.phyloP100way.wigFix.gz done # real 55m10.637s du -hsc downloads # 5.3G downloads # check integrity of data with wigToBigWig time (zcat downloads/*.wigFix.gz \ | wigToBigWig -verbose=2 stdin /hive/data/genomes/hg38/chrom.sizes \ phyloP100way.bw) > bigWig.log 2>&1 egrep "real|VmPeak" bigWig.log # pid=66292: VmPeak: 33751268 kB # real 43m40.194s bigWigInfo phyloP100way.bw | sed -e 's/^/# /;' # version: 4 # isCompressed: yes # isSwapped: 0 # primaryDataSize: 7,631,142,234 # primaryIndexSize: 93,060,056 # zoomLevels: 10 # chromCount: 356 # basesCovered: 2,944,991,967 # mean: 0.093059 # min: -20.000000 # max: 10.003000 # std: 1.036944 # encode those files into wiggle data time (zcat downloads/*.wigFix.gz \ | wigEncode stdin phyloP100way.wig phyloP100way.wib) # Converted stdin, upper limit 10.00, lower limit -20.00 # real 16m50.606s du -hsc *.wi? # 2.8G phyloP100way.wib # 293M phyloP100way.wig # Load gbdb and database with wiggle. ln -s `pwd`/phyloP100way.wib /gbdb/hg38/multiz100way/phyloP100way.wib time hgLoadWiggle -pathPrefix=/gbdb/hg38/multiz100way hg38 \ phyloP100way phyloP100way.wig # real 0m32.623s # use to set trackDb.ra entries for wiggle min and max # and verify table is loaded correctly wigTableStats.sh hg38 phyloP100way # db.table min max mean count sumData # hg38.phyloP100way -20 10.003 0.0930586 2944991967 2.74057e+08 # stdDev viewLimits # 1.03694 viewLimits=-5.09166:5.27778 # that range is: 20+10.003 = 30.003 for hBinSize=0.030003 # Create histogram to get an overview of all the data time hgWiggle -doHistogram \ -hBinSize=0.030003 -hBinCount=1000 -hMinVal=-20 -verbose=2 \ -db=hg38 phyloP100way > histogram.data 2>&1 # real 2m42.342s # find out the range for the 2:5 graph grep -v chrom histogram.data | grep "^[0-9]" | ave -col=5 stdin \ | sed -e 's/^/# /;' # Q1 0.000000 # median 0.000023 # Q3 0.000186 # average 0.001145 # min 0.000000 # max 0.035821 # count 873 # total 0.999972 # standard deviation 0.003928 # create plot of histogram: cat << '_EOF_' | gnuplot > histo.png set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff font \ "/usr/share/fonts/default/Type1/n022004l.pfb" set size 1.4, 0.8 set key left box set grid noxtics set grid ytics set title " Human hg38 Histogram phyloP100way track" set xlabel " phyloP100way score" set ylabel " Relative Frequency" set y2label " Cumulative Relative Frequency (CRF)" set y2range [0:1] set y2tics set xrange [-3:3] set yrange [0:0.04] plot "histogram.data" using 2:5 title " RelFreq" with impulses, \ "histogram.data" using 2:7 axes x1y2 title " CRF" with lines '_EOF_' # << happy emacs set xrange [-4:1] display histo.png & ############################################################################# # construct download files for 100-way (DONE - 2015-05-12 - Hiram) mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/multiz100way mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/phastCons100way mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/phyloP100way mkdir /hive/data/genomes/hg38/bed/multiz100way/downloads cd /hive/data/genomes/hg38/bed/multiz100way/downloads mkdir multiz100way phastCons100way phyloP100way cd multiz100way mkdir maf rsync -a -P ../../anno/result/ ./maf/ du -hsc maf/ # 789G maf cd maf time gzip *.maf & # real 476m2.754s du -hsc maf # 72G maf cd maf md5sum *.gz > md5sum.txt mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/multiz100way/maf ln -s `pwd`/* /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/multiz100way/maf ln -s ../../hg38.100way.commonNames.nh . ln -s ../../hg38.100way.scientificNames.nh . ~/kent/src/hg/utils/phyloTrees/asciiTree.pl ../../hg38.100way.nh \ > hg38.100way.nh # obtain the README.txt from hg38/multiz7way and update for this # situation time md5sum *.nh *.maf.gz > md5sum.txt # real 1m55.3100s ln -s `pwd`/*.txt `pwd`/*.nh `pwd`/*.gz \ /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/multiz100way ##################################################################### cd /hive/data/genomes/hg38/bed/multiz100way/downloads/phastCons100way mkdir hg38.100way.phastCons cd hg38.100way.phastCons ln -s ../../../cons/all/downloads/*.wigFix.gz . md5sum *.gz > md5sum.txt cd .. ln -s ../../cons/all/phastCons100way.bw ./hg38.phastCons100way.bw ln -s ../../cons/all/all.mod ./hg38.phastCons100way.mod time md5sum *.mod *.bw > md5sum.txt # real 0m47.434s # obtain the README.txt from hg19/phastCons100way and update for this # situation mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/phastCons100way/hg38.100way.phastCons cd hg38.100way.phastCons ln -s `pwd`/* /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/phastCons100way/hg38.100way.phastCons cd .. ln -s `pwd`/*.mod `pwd`/*.bw `pwd`/*.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/phastCons100way ##################################################################### cd /hive/data/genomes/hg38/bed/multiz100way/downloads/phyloP100way mkdir hg38.100way.phyloP100way cd hg38.100way.phyloP100way ln -s ../../../consPhyloP/all/downloads/*.gz . time md5sum *.gz > md5sum.txt # real 0m25.736s ln -s ../../consPhyloP/run.phyloP/all.mod hg38.phyloP100way.mod ln -s ../../consPhyloP/all/phyloP100way.bw hg38.phyloP100way.bw # obtain the README.txt from hg38/phyloP7way and update for this # situation time md5sum *.mod *.bw *.txt > md5sum.txt # real 0m34.249s ln -s `pwd`/*.bw `pwd`/*.mod `pwd`/*.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/phyloP100way ########################################################################### ## create upstream refGene maf files cd /hive/data/genomes/hg38/bed/multiz100way/downloads/multiz100way # bash script #!/bin/sh export geneTbl="knownGene" for S in 1000 2000 5000 do echo "making upstream${S}.maf" featureBits hg38 ${geneTbl}:upstream:${S} -fa=/dev/null -bed=stdout \ | perl -wpe 's/_up[^\t]+/\t0/' | sort -k1,1 -k2,2n \ | /cluster/bin/$MACHTYPE/mafFrags hg38 multiz100way \ stdin stdout \ -orgs=/hive/data/genomes/hg38/bed/multiz100way/species.list \ | gzip -c > upstream${S}.${geneTbl}.maf.gz echo "done upstream${S}.${geneTbl}.maf.gz" done # real 384m21.270s md5sum upstream*.gz >> md5sum.txt # obtain the README.txt from hg38/multiz7way and update for this # situation # information for table of species in the README files, need to # edit it in after adding it to the end of this file: cat ../../species.list | tr '[ ]' '[\n]' | while read D do netType=`ls ../../mafLinks/${D}/hg38.${D}.*.maf.gz | sed -e "s#.*hg38.${D}.##; s#.maf.gz##;" | sed -e 's/synNet/syntenic/; s/rbest/reciprocal best/;'` info=`hgsql -N -e "select organism,\" - \",scientificName,description from dbDb where name=\"$D\";" hgcentraltest` echo "${info} ${netType}" done | tr '[\t]' '[ ]' >> README.txt # some other symlinks were already made above ln -s `pwd`/upstream*.gz README.txt \ /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/multiz100way ############################################################################# # hgPal downloads (rebuilt knownGene and knownCanonical 2016-06-27 braney) ssh hgwdev screen -S hg38HgPal mkdir /hive/data/genomes/hg38/bed/multiz100way/pal cd /hive/data/genomes/hg38/bed/multiz100way/pal cat ../species.list | tr '[ ]' '[\n]' > order.list export mz=multiz100way export gp=knownGene export db=hg38 export I=0 mkdir exonAA exonNuc for C in `sort -nk2 ../../../chrom.sizes | cut -f1` do I=`echo $I | awk '{print $1+1}'` echo "mafGene -chrom=$C -exons -noTrans $db $mz $gp order.list stdout | gzip -c > exonNuc/$C.exonNuc.fa.gz &" echo "mafGene -chrom=$C -exons $db $mz $gp order.list stdout | gzip -c > exonAA/$C.exonAA.fa.gz &" if [ $I -gt 11 ]; then echo "date" echo "wait" I=0 fi done > $gp.jobs echo "date" >> $gp.jobs echo "wait" >> $gp.jobs time sh -x ./$gp.jobs > $gp.jobs.log 2>&1 & # real 208m39.304s time cat exonAA/*.gz > $gp.$mz.exonAA.fa.gz # real 0m6.023s time cat exonNuc/*.gz > $gp.$mz.exonNuc.fa.gz # real 0m9.152s export mz=multiz100way export gp=knownGene export db=hg38 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/ rm -rf exonAA exonNuc ### need other gene track alignments also # running up refGene cd /hive/data/genomes/hg38/bed/multiz100way/pal export mz=multiz100way export gp=ncbiRefSeq export db=hg38 export I=0 mkdir exonAA exonNuc for C in `sort -nk2 ../../../chrom.sizes | cut -f1` do I=`echo $I | awk '{print $1+1}'` echo "mafGene -chrom=$C -exons -noTrans $db $mz $gp order.list stdout | gzip -c > exonNuc/$C.exonNuc.fa.gz &" echo "mafGene -chrom=$C -exons $db $mz $gp order.list stdout | gzip -c > exonAA/$C.exonAA.fa.gz &" if [ $I -gt 11 ]; then echo "date" echo "wait" I=0 fi done > $gp.jobs echo "date" >> $gp.jobs echo "wait" >> $gp.jobs time sh -x $gp.jobs > $gp.jobs.log 2>&1 # real 126m0.688s export mz=multiz100way export gp=ncbiRefSeq export db=hg38 time cat exonAA/*.gz > $gp.$mz.exonAA.fa.gz # real 3m14.449s time cat exonNuc/*.gz> $gp.$mz.exonNuc.fa.gz # real 13m27.577s du -hsc exonAA exonNuc $gp*.fa.gz # 3.1G exonAA # 4.9G exonNuc # 3.1G ncbiRefSeq.multiz100way.exonAA.fa.gz # 4.9G ncbiRefSeq.multiz100way.exonNuc.fa.gz rm -rf exonAA exonNuc # we're only distributing exons at the moment export mz=multiz100way export gp=ncbiRefSeq export db=hg38 export pd=/usr/local/apache/htdocs-hgdownload/goldenPath/$db/$mz/alignments mkdir -p $pd md5sum $gp.*.fa.gz >> md5sum.txt ln -s `pwd`/$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/ ### And knownCanonical cd /hive/data/genomes/hg38/bed/multiz100way/pal export mz=multiz100way export gp=knownCanonical export db=hg38 mkdir exonAA exonNuc knownCanonical time cut -f1 ../../../chrom.sizes | while read C do echo $C 1>&2 hgsql hg38 -N -e "select chrom, chromStart, chromEnd, transcript from knownCanonical where chrom='$C'" > knownCanonical/$C.known.bed done # real 0m15.897s ls knownCanonical/*.known.bed | while read F do if [ -s $F ]; then echo $F | sed -e 's#knownCanonical/##; s/.known.bed//' fi done | while read C do echo "date" echo "mafGene -geneBeds=knownCanonical/$C.known.bed -exons -noTrans $db $mz knownGene order.list stdout | \ gzip -c > exonNuc/$C.exonNuc.fa.gz " echo "mafGene -geneBeds=knownCanonical/$C.known.bed -exons $db $mz knownGene order.list stdout | \ gzip -c > exonAA/$C.exonAA.fa.gz " done > $gp.$mz.jobs time sh -x $gp.$mz.jobs > $gp.$mz.job.log 2>&1 # 109m16.821s rm *.known.bed export mz=multiz100way export gp=knownCanonical export db=hg38 cat exonAA/c*.gz > $gp.$mz.exonAA.fa.gz cat exonNuc/c*.gz > $gp.$mz.exonNuc.fa.gz + + mkdir protAA protNuc + + ls knownCanonical/*.known.bed | while read F + do + if [ -s $F ]; then + echo $F | sed -e 's#knownCanonical/##; s/.known.bed//' + fi + done | while read C + do + echo "date" + echo "mafGene -geneBeds=knownCanonical/$C.known.bed -noTrans $db $mz knownGene order.list stdout | \ + gzip -c > protNuc/$C.protNuc.fa.gz" + echo "mafGene -geneBeds=knownCanonical/$C.known.bed $db $mz knownGene order.list stdout | \ + gzip -c > protAA/$C.protAA.fa.gz" + done > $gp.$mz.prot.jobs + + time (sh -x $gp.$mz.jobs) > $gp.$mz.prot.job.log + # real 230m33.479s + + export mz=multiz100way + export gp=knownCanonical + export db=hg38 + zcat protAA/c*.gz | gzip -c > $gp.$mz.protAA.fa.gz & + zcat protNuc/c*.gz | gzip -c > $gp.$mz.protNuc.fa.gz & + # about 6 minutes + rm -rf exonAA exonNuc export mz=multiz100way export gp=knownCanonical export db=hg38 export pd=/usr/local/apache/htdocs-hgdownload/goldenPath/$db/$mz/alignments mkdir -p $pd ln -s `pwd`/$gp.$mz.exonAA.fa.gz $pd/$gp.exonAA.fa.gz ln -s `pwd`/$gp.$mz.exonNuc.fa.gz $pd/$gp.exonNuc.fa.gz cd $pd md5sum *.fa.gz > md5sum.txt ############################################################################# # wiki page for 100-way (DONE - 2015-05-14 - Hiram) mkdir /hive/users/hiram/bigWays/hg38.100way cd /hive/users/hiram/bigWays echo "hg38" > hg38.100way/ordered.list awk '{print $1}' /hive/data/genomes/hg38/bed/multiz100way/100way.distances.txt \ >> hg38.100way/ordered.list # sizeStats.sh catches up the cached measurements required for data # in the tables. They may already be done. ./sizeStats.sh hg38.100way/ordered.list # dbDb.sh constructs hg38.100way/Hg38_100-way_conservation_alignment.html ./dbDb.sh hg38 100way # sizeStats.pl constructs hg38.100way/Hg38_100-way_Genome_size_statistics.html ./sizeStats.pl hg38 100way # defCheck.pl constructs Hg38_100-way_conservation_lastz_parameters.html ./defCheck.pl hg38 100way # this constructs the html pages in hg38.100way/: # -rw-rw-r-- 1 53064 May 14 12:47 Hg38_100-way_conservation_alignment.html # -rw-rw-r-- 1 65315 May 14 12:49 Hg38_100-way_Genome_size_statistics.html # -rw-rw-r-- 1 31588 May 14 12:49 Hg38_100-way_conservation_lastz_parameters.html # add those pages to the genomewiki. Their page names are the # names of the .html files without the .html: # Hg38_100-way_conservation_alignment # Hg38_100-way_Genome_size_statistics # Hg38_100-way_conservation_lastz_parameters # when you view the first one you enter, it will have links to the # missing two. ############################################################################# # make default species maf (DONE braney 2016-09-28) cd /cluster/data/hg38/bed/multiz100way/anno mkdir defaultSpecies cd defaultSpecies for i in ../result/*.maf; do echo $i; mafSpeciesSubset $i list.txt `basename $i`; done mkdir /gbdb/hg38/multiz100way/defaultMaf for i in *.maf; do ln -s `pwd`/$i.maf /gbdb/hg38/multiz100way/defaultMaf; done hgLoadMaf hg38 mutliz100wayDefault -pathPrefix=/gbdb/hg38/multiz100way/defaultMaf