282d648d8896a7bb45af20bbc306873f6a5ffb87
angie
Wed Jan 17 11:03:32 2024 -0800
Added chrUn_KI270752v1 to hg38.{p11,p12,p13}.chromAlias.txt files after user pointed out in MLQ #32874 that it was missing.
It is present in hg38.p14.chromAlias.txt and all *chromAlias.bb files.
diff --git src/hg/makeDb/doc/hg38/patchUpdate.11.txt src/hg/makeDb/doc/hg38/patchUpdate.11.txt
index 28285ff..028c090 100644
--- src/hg/makeDb/doc/hg38/patchUpdate.11.txt
+++ src/hg/makeDb/doc/hg38/patchUpdate.11.txt
@@ -1,700 +1,721 @@
# for emacs: -*- mode: sh; -*-
# This file describes how hg38 was extended with patch sequences and annotations from grcH38P11
# Hold off on actually installing these until the genbank process has produced tables on hg38
##############################################################################
# Extend main database 2bit, chrom.sizes, chromInfo (DONE - 2018-08-06 - Angie)
cd /hive/data/genomes/hg38
# main 2bit
time faToTwoBit <(twoBitToFa hg38.2bit stdout) \
<(twoBitToFa /hive/data/genomes/grcH38P11/grcH38P11.2bit stdout) \
hg38.p11.2bit
#real 1m30.733s
# unmasked 2bit
twoBitMask -type=.bed hg38.p11.2bit /dev/null hg38.p11.unmasked.2bit
# chrom.sizes
sort -k2nr,2nr chrom.sizes /hive/data/genomes/grcH38P11/chrom.sizes > chrom.sizes.p11
# chromInfo
cd /hive/data/genomes/hg38/bed/chromInfo
awk '{print $1 "\t" $2 "\t/gbdb/hg38/hg38.2bit";}' ../../chrom.sizes.p11 \
> chromInfo.p11.tab
wc -l chromInfo*.tab
# 578 chromInfo.p11.tab
# 455 chromInfo.tab
# Install
cd /hive/data/genomes/hg38
# For the first update only, move initial release files to .initial. Don't do this next update!
mv hg38.2bit hg38.initial.2bit
ln -s hg38.p11.2bit hg38.2bit
mv hg38.unmasked.2bit hg38.initial.unmasked.2bit
ln -s hg38.p11.unmasked.2bit hg38.unmasked.2bit
mv chrom.sizes chrom.sizes.initial
ln -s chrom.sizes.p11 chrom.sizes
cd /hive/data/genomes/hg38/bed/chromInfo
hgLoadSqlTab hg38 chromInfo chromInfo.sql chromInfo.p11.tab
##############################################################################
# Extend main database tables for fileless tracks (DONE - 2018-08-06 - Angie)
# Just add the patch table rows to the main database tables
for table in gap gold rmsk simpleRepeat windowmaskerSdust cpgIslandExt genscan augustusGene; do
echo $table
hgsql hg38 -e "insert into hg38.$table select * from grcH38P11.$table"
done
##############################################################################
# Extend main database gc5BaseBw.bw (DONE - 2018-08-06 - Angie)
cd /hive/data/genomes/hg38/bed/gc5Base/
# Concatenate original assembly results with grcH38P11 results
time (zcat hg38.gc5Base.wigVarStep.gz \
/hive/data/genomes/grcH38P11/bed/gc5Base/grcH38P11.gc5Base.wigVarStep.gz \
| gzip -c \
> hg38.p11.gc5Base.wigVarStep.gz)
#real 8m10.061s
# Make a new gc5BaseBw.bw
time wigToBigWig hg38.p11.gc5Base.wigVarStep.gz ../../chrom.sizes.p11 \
hg38.p11.gc5Base.bw
#real 16m28.272s
# Install
cd /hive/data/genomes/hg38/bed/gc5Base/
# For the first update only, move initial release files to .initial. Don't do this next update!
mv hg38.gc5Base.wigVarStep.gz hg38.initial.gc5Base.wigVarStep.gz
mv hg38.gc5Base.bw hg38.initial.gc5Base.bw
# End of first-update-only stuff
ln -s hg38.p11.gc5Base.wigVarStep.gz hg38.gc5Base.wigVarStep.gz
ln -s hg38.p11.gc5Base.bw hg38.gc5Base.bw
# Because of this symlink, browser track gc5BaseBw has been automatically updated:
# /gbdb/hg38/bbi/gc5BaseBw/gc5Base.bw -> /cluster/data/hg38/bed/gc5Base/hg38.gc5Base.bw
# For the first update only. Don't do this next update!
# The .wib and .wig.gz are obsolete, remove them from the downloads.
rm /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/bigZips/hg38.gc5Base.wib
rm /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/bigZips/hg38.gc5Base.wig.gz
# link the bigwig to downloads
ln -sf /hive/data/genomes/hg38/bed/gc5Base/hg38.gc5Base.bw /usr/local/apache/htdocs-hgdownload/goldenPath/hg38/bigZips/hg38.gc5Base.bw
# End of first-update-only stuff
##############################################################################
# Extend main database download files (DONE - 2018-08-06 - Angie)
cd /hive/data/genomes/hg38/goldenPath/bigZips
mkdir p11
# hg38.2bit was already extended above.
ln -sf /hive/data/genomes/hg38/hg38.p11.2bit p11/
# AGP:
zcat hg38.agp.gz \
/hive/data/genomes/grcH38P11/goldenPath/bigZips/grcH38P11.agp.gz \
| grep -v ^# \
| gzip -c > p11/hg38.p11.agp.gz
# FASTA
twoBitToFa ../../hg38.p11.2bit stdout \
| gzip -c > p11/hg38.p11.fa.gz
twoBitToFa hg38.2bit stdout \
| maskOutFa stdin hard stdout \
| gzip -c > p11/hg38.p11.fa.masked.gz
# RepeatMasker (don't include header of patch file):
cat <(zcat hg38.fa.out.gz) \
<(zcat /hive/data/genomes/grcH38P11/goldenPath/bigZips/grcH38P11.fa.out.gz | tail -n +4) \
| gzip -c > p11/hg38.p11.fa.out.gz
# SimpleRepeats/TRF:
zcat hg38.trf.bed.gz \
/hive/data/genomes/grcH38P11/goldenPath/bigZips/grcH38P11.trf.bed.gz \
| gzip -c > p11/hg38.p11.trf.bed.gz
# We don't expect a complete set of chroms to have simpleRepeats, but at least an increase:
zcat p11/hg38.trf.bed.gz | cut -f 1 | uniq | wc -l
#363
zcat p11/hg38.p11.trf.bed.gz | cut -f 1 | uniq | wc -l
#485
# hg38 files that are not built by makeDownloads.pl because hg38 is treated as 'scaffold-based':
# Per-chrom soft-masked FASTA:
rm -rf chroms
tar xvzf hg38.chromFa.tar.gz
faSplit byname /hive/data/genomes/grcH38P11/goldenPath/bigZips/grcH38P11.fa.gz chroms/
ls -1 chroms | wc -l
#578
tar cvzf p11/hg38.p11.chromFa.tar.gz ./chroms
rm -rf chroms
# Per-chrom hard-masked FASTA:
rm -rf maskedChroms
tar xvzf hg38.chromFaMasked.tar.gz
faSplit byname /hive/data/genomes/grcH38P11/goldenPath/bigZips/grcH38P11.fa.masked.gz \
maskedChroms/
ls -1 maskedChroms | wc -l
#578
tar cvzf p11/hg38.p11.chromFaMasked.tar.gz ./maskedChroms
rm -rf maskedChroms
# RepeatMasker .align files:
zcat hg38.fa.align.gz /hive/data/genomes/grcH38P11/bed/repeatMasker/grcH38P11.fa.align.gz \
| gzip -c > p11/hg38.p11.fa.align.gz
# Make new md5sum.txt
cd p11
md5sum hg38.* > md5sum.txt
# Install
cd /hive/data/genomes/hg38/goldenPath/bigZips
mkdir initial
# For the first update only, move initial release files to initial/. Don't do this next update!
mv hg38.* md5sum.txt initial/
ln -s p11/README.txt .
for file in p11/hg38.p11* p11/md5sum.txt; do
linkName=$(echo $file | sed -e 's/p11.//g')
ln -sf $file $linkName
done
# First time only: update symlinks used for initial downloads.
# Won't be necessary for subsequent patches.
ln -sf /hive/data/genomes/hg38/hg38.initial.2bit initial/hg38.2bit
ln -sf /hive/data/genomes/hg38/bed/rmskCM/hg38.fa.align.gz initial/hg38.fa.align.gz
# Edit README.txt
#############################################################################
# Extend cytoBand{,Ideo} (DONE 18-08-06 angie)
cd /hive/data/genomes/hg38/bed/cytoBand
tawk '{print $1, 0, $2, "", "gneg";}' /hive/data/genomes/grcH38P11/chrom.sizes \
> cytoBand.p11.tab
hgLoadSqlTab -oldTable hg38 cytoBand - cytoBand.p11.tab
hgLoadSqlTab -oldTable hg38 cytoBandIdeo - cytoBand.p11.tab
#########################################################################
# Regenerate idKeys with extended hg38 (DONE - 2018-08-06 - Angie)
mkdir /hive/data/genomes/hg38/bed/idKeys.p11
cd /hive/data/genomes/hg38/bed/idKeys.p11
# ku down... use hgwdev this time:
time ($HOME/kent/src/hg/utils/automation/doIdKeys.pl \
-twoBit=/hive/data/genomes/hg38/hg38.p11.unmasked.2bit \
-bigClusterHub=hgwdev -smallClusterHub=hgwdev \
-buildDir=`pwd` hg38) > do.log 2>&1 &
tail -f do.log
#real 1m25.689s
cat hg38.keySignature.txt
#15f8c2af14b6aaaef08775dbf0c8e900
# Install
# For the first update only, move initial release files to .initial. Don't do this next update!
mv /hive/data/genomes/hg38/bed/idKeys{,.initial}
cd /hive/data/genomes/hg38/bed/
ln -s idKeys.p11 idKeys
##############################################################################
# UCSC to RefSeq, INSDC, Assembly; chromAlias (DONE 18-08-07 angie)
# need to have idKeys for the genbank and refseq assemblies:
mkdir -p /hive/data/genomes/hg38/bed/ucscToINSDC/genbankP11
cd /hive/data/genomes/hg38/bed/ucscToINSDC/genbankP11
ln -s /hive/data/outside/ncbi/genomes/genbank/vertebrate_mammalian/Homo_sapiens/all_assembly_versions/GCA_000001405.26_GRCh38.p11/GCA_000001405.26_GRCh38.p11_genomic.fna.gz .
faToTwoBit GCA_000001405.26_GRCh38.p11_genomic.fna.gz genbankP11.2bit
time ($HOME/kent/src/hg/utils/automation/doIdKeys.pl -buildDir=`pwd` -twoBit=genbankP11.2bit \
-bigClusterHub=hgwdev -smallClusterHub=hgwdev \
genbankP11) > do.log 2>&1
#real 1m50.109s
mkdir /hive/data/genomes/hg38/bed/ucscToINSDC/refseqP11
cd /hive/data/genomes/hg38/bed/ucscToINSDC/refseqP11
ln -s /hive/data/outside/ncbi/genomes/refseq/vertebrate_mammalian/Homo_sapiens/all_assembly_versions/GCF_000001405.37_GRCh38.p11/GCF_000001405.37_GRCh38.p11_genomic.fna.gz ./
faToTwoBit GCF_000001405.37_GRCh38.p11_genomic.fna.gz refseqP11.2bit
time ($HOME/kent/src/hg/utils/automation/doIdKeys.pl -buildDir=`pwd` -twoBit=refseqP11.2bit \
-bigClusterHub=hgwdev -smallClusterHub=hgwdev \
refseqP11) > do.log 2>&1
#real 1m47.878s
# with the three idKeys available, join them to make the table bed files:
cd /hive/data/genomes/hg38/bed/ucscToINSDC
join -t$'\t' ../idKeys/hg38.idKeys.txt genbankP11/genbankP11.idKeys.txt \
| cut -f2- | sort -k1,1 | join -t$'\t' <(sort -k1,1 ../../chrom.sizes) - \
| awk '{printf "%s\t0\t%d\t%s\n", $1, $2, $3}' \
| sort -k1,1 -k2,2n > ucscToINSDC.p11.bed
join -t$'\t' ../idKeys/hg38.idKeys.txt refseqP11/refseqP11.idKeys.txt \
| cut -f2- | sort -k1,1 | join -t$'\t' <(sort -k1,1 ../../chrom.sizes) - \
| awk '{printf "%s\t0\t%d\t%s\n", $1, $2, $3}' \
| sort -k1,1 -k2,2n > ucscToRefSeq.p11.bed
# loading tables:
export db=hg38
export chrSize=`cut -f1 ucscToINSDC.p11.bed | awk '{print length($0)}' | sort -n | tail -1`
sed -e "s/21/$chrSize/" $HOME/kent/src/hg/lib/ucscToINSDC.sql \
| hgLoadSqlTab ${db} ucscToINSDC stdin ucscToINSDC.p11.bed
export chrSize=`cut -f1 ucscToRefSeq.p11.bed | awk '{print length($0)}' | sort -n | tail -1`
sed -e "s/21/$chrSize/" $HOME/kent/src/hg/lib/ucscToINSDC.sql \
| sed -e 's/INSDC/RefSeq/g;' \
| hgLoadSqlTab ${db} ucscToRefSeq stdin ucscToRefSeq.p11.bed
# must be exactly 100% coverage
featureBits -countGaps ${db} ucscToINSDC
#3257347282 bases of 3257347282 (100.000%) in intersection
featureBits -countGaps ${db} ucscToRefSeq
#3257319537 bases of 3257347282 (99.999%) in intersection
# construct chromAlias:
cd /hive/data/genomes/hg38/bed/chromAlias
hgsql -N -e 'select chrom,name from ucscToRefSeq;' ${db} \
| sort -k1,1 > ucsc.refseq.tab
hgsql -N -e 'select chrom,name from ucscToINSDC;' ${db} \
| sort -k1,1 > ucsc.genbank.tab
# add NCBI sequence names from assembly report
grep -v ^# \
/hive/data/genomes/grcH38P11/genbank/GCA_000001405.26_GRCh38.p11_assembly_report.txt \
| tawk '{print $5, $1;}' | sort \
> genbankToAssembly.txt
tawk '{print $2, $1;}' ucsc.genbank.tab | sort \
| join -t$'\t' -o 1.2,2.2 - genbankToAssembly.txt \
| sort -k1,1 > ucsc.assembly.tab
~/kent/src/hg/utils/automation/chromAlias.pl ucsc.*.tab \
> ${db}.chromAlias.tab
# verify all there:
for t in refseq genbank assembly
do
c0=`cat ucsc.$t.tab | wc -l`
c1=`grep $t hg38.chromAlias.tab | wc -l`
ok="OK"
if [ "$c0" -ne "$c1" ]; then
ok="ERROR"
fi
printf "# checking $t: $c0 =? $c1 $ok\n"
done
# checking refseq: 578 =? 578 OK
# checking genbank: 578 =? 578 OK
# checking assembly: 578 =? 578 OK
hgLoadSqlTab hg38 chromAlias $HOME/kent/src/hg/lib/chromAlias.sql ${db}.chromAlias.tab
##############################################################################
# UCSC to Ensembl (TODO 18-08-06 angie)
# doc??
############################################################################
# altLocations and patchLocations (DONE - 2018-08-06 - Angie)
# indicate corresponding locations between haplotypes and reference
mkdir /hive/data/genomes/hg38/bed/altLocations.p11
cd /hive/data/genomes/hg38/bed/altLocations.p11
~/kent/src/hg/utils/automation/altScaffoldPlacementToBed.pl \
/hive/data/genomes/grcH38P11/genbank/GCA_000001405.26_GRCh38.p11_assembly_structure/{ALT_*,PATCHES}/alt_scaffolds/alt_scaffold_placement.txt \
| sort -k1,1 -k2n,2n \
> altAndFixLocations.bed
wc -l altAndFixLocations.bed
#768 altAndFixLocations.bed
grep _alt altAndFixLocations.bed > altLocations.bed
grep _fix altAndFixLocations.bed > fixLocations.bed
hgLoadBed hg38 altLocations{,.bed}
#Read 642 elements of size 4 from altLocations.bed
hgLoadBed hg38 fixLocations{,.bed}
#Read 128 elements of size 4 from fixLocations.bed
featureBits -countGaps hg38 altLocations
#196222738 bases of 3253848404 (6.030%) in intersection
featureBits -countGaps hg38 fixLocations
#60769916 bases of 3253848404 (1.868%) in intersection
#############################################################################
# Check for new chrX alts/patches to add to par (DONE 2018-08-06 angie)
# Thanks to Hiram for pointing out that intersecting chrX positions in
# altLocations and par shows whether a chrX alt overlaps a PAR.
cd /hive/data/genomes/hg38/bed/par
hgsql hg38 -e 'select * from altLocations where chrom like "chrX%"'
#+-----+---------------------+------------+----------+------------------------+
#| bin | chrom | chromStart | chromEnd | name |
#+-----+---------------------+------------+----------+------------------------+
#| 73 | chrX | 319337 | 601516 | chrX_KI270880v1_alt |
#| 73 | chrX | 326487 | 601516 | chrX_KI270913v1_alt |
#| 149 | chrX | 79965153 | 80097082 | chrX_KI270881v1_alt |
#| 73 | chrX_KI270880v1_alt | 0 | 284869 | chrX:319338-601516 |
#| 73 | chrX_KI270881v1_alt | 0 | 144206 | chrX:79965154-80097082 |
#| 73 | chrX_KI270913v1_alt | 0 | 274009 | chrX:326488-601516 |
#+-----+---------------------+------------+----------+------------------------+
hgsql hg38 -e 'select * from par where chrom like "chrX%"'
#+-----+---------------------+------------+-----------+------+
#| bin | chrom | chromStart | chromEnd | name |
#+-----+---------------------+------------+-----------+------+
#| 9 | chrX | 10000 | 2781479 | PAR1 |
#| 221 | chrX | 155701382 | 156030895 | PAR2 |
#| 73 | chrX_KI270880v1_alt | 0 | 284869 | PAR1 |
#| 73 | chrX_KI270913v1_alt | 0 | 274009 | PAR1 |
#+-----+-------+------------+-----------+------+
# chrX_KI270881v1_alt is not in either PAR.
# chrX_KI270880v1_alt and chrX_KI270913v1_alt are entirely contained in PAR1 --
# and are already in the PAR table, so nothing to add.
##############################################################################
# altSeqLiftOver (DONE 19-01-07 Angie)
# originally done 18-08-07; redone 18-11-06 with fixed gff3ToPsl to get correct - strand alignments
# mainToPatch over.chain regenerated 19-01-07 with fixed pslToChain
mkdir /hive/data/genomes/hg38/bed/altSeqLiftOver.p11
cd /hive/data/genomes/hg38/bed/altSeqLiftOver.p11
# Eventually these will be under the /hive/data/genomes/.../genbank/... directory
# that points to /hive/data/outside/ncbi/genomes/... but at the moment the contents
# of the alignments/ directories are not included in the sync. So for now,
# manually download them here.
# Original alts:
mkdir initialAlts
cd initialAlts
for d in /hive/data/genomes/grcH38P11/genbank/GCA_000001405.26_GRCh38.p11_assembly_structure/ALT*/alt_scaffolds/alignments; do
subdir=$(echo $d | sed -re 's@^/hive/data/genomes/grcH38P11/genbank/@@;')
wget --timestamping --no-verbose \
ftp://ftp.ncbi.nlm.nih.gov/genomes/genbank/vertebrate_mammalian/Homo_sapiens/all_assembly_versions/GCA_000001405.26_GRCh38.p11/$subdir/\*.gff
done
# New alts and patches too:
mkdir ../patches
cd ../patches
wget --timestamping --no-verbose\
ftp://ftp.ncbi.nlm.nih.gov/genomes/genbank/vertebrate_mammalian/Homo_sapiens/all_assembly_versions/GCA_000001405.26_GRCh38.p11/GCA_000001405.26_GRCh38.p11_assembly_structure/PATCHES/alt_scaffolds/alignments/\*.gff
cd ..
# Use chromAlias to make a .sed file to substitute Genbank accessions to UCSC names
hgsql hg38 -NBe 'select alias,chrom from chromAlias where find_in_set("genbank", source);' \
| awk '{print "s@" $1 "@" $2 "@;";}' > gbToUcsc.sed
cp /dev/null altToChrom.noScore.psl
for f in initialAlts/*.gff patches/*.gff; do
e=$(basename $f .gff | sed -e 's/_/|/g;')
s=$(grep -E $e gbToUcsc.sed)
sed -re "$s" $f | gff3ToPsl ../../chrom.sizes{,} stdin stdout \
| pslPosTarget stdin stdout \
>> altToChrom.noScore.psl
done
pslCheck altToChrom.noScore.psl
#checked: 404 failed: 0 errors: 0
time pslRecalcMatch altToChrom.noScore.psl ../../hg38.2bit{,} altToChrom.psl
#202.461u 1.836s 3:24.46 99.9% 0+0k 0+0io 0pf+0w
pslSwap altToChrom.psl stdout | pslPosTarget stdin chromToAlt.psl
sort -k14,14 -k16n,16n -k10,10 -k12n,12n altToChrom.psl chromToAlt.psl \
> altAndPatches.psl
grep _alt altAndPatches.psl > altSeqLiftOver.psl
grep _fix altAndPatches.psl > fixSeqLiftOver.psl
# Load tables
# Not reloaded 18-11-06 because tables have been reloaded with alignments through p12,
# see patchUpdate.12.txt
hgLoadPsl hg38 -table=altSeqLiftOverPsl altSeqLiftOver.psl
hgLoadPsl hg38 -table=fixSeqLiftOverPsl fixSeqLiftOver.psl
# Make chrom-to-alt PSL file for genbank process.
ln -f -s `pwd`/chromToAlt.psl \
/hive/data/genomes/hg38/jkStuff/hg38.p11.alt.psl
# Make a liftOver chain file for mapping annotations on main chroms to new patch sequences
# 6/15/18: exclude alts that were already in hg38 before p11
# Redone 1/7/19 after Braney fixed pslToChain
cut -f 1 ../../chrom.sizes.initial | grep _ \
| grep -vwf - chromToAlt.psl \
| pslToChain stdin stdout \
| chainScore stdin ../../hg38.2bit{,} ../../jkStuff/hg38.mainToPatch.p11.over.chain
#52.068u 1.626s 0:54.43 98.6% 0+0k 15952+0io 2pf+0w
# 1/7/19 also make a liftOver that includes the original alts, for tracks that have
# annotations only on main chromosomes. Exclude alt-to-fix alignments.
# This is necessary only for the first time we add a patch update.
awk '($14 !~ /_/)' chromToAlt.psl \
| pslToChain stdin stdout \
| chainScore stdin ../../hg38.2bit{,} ../../jkStuff/hg38.mainToAllAltPatch.p11.over.chain
#23.971u 1.400s 0:25.34 100.1% 0+0k 0+0io 0pf+0w
#########################################################################
# ncbiRefSeq.p11 Genes (DONE - 2018-08-08 - Angie)
mkdir /hive/data/genomes/hg38/bed/ncbiRefSeq.p11.2018-08-08
cd /hive/data/genomes/hg38/bed/ncbiRefSeq.p11.2018-08-08
time (~/kent/src/hg/utils/automation/doNcbiRefSeq.pl -buildDir=`pwd` \
refseq vertebrate_mammalian Homo_sapiens \
GCF_000001405.37_GRCh38.p11 hg38) > do.log 2>&1 & tail -f do.log
# *** All done ! Elapsed time: 15m24s
#real 15m24.236s
cat fb.ncbiRefSeq.hg38.txt
#131634821 bases of 3092500061 (4.257%) in intersection
##############################################################################
# Extend wgEncodeReg bigWig tracks (DONE 19-01-07 angie)
# first done 18-08-10; redone 18-11-06 and 19-01-07 with updated .over.chain file.
# NOTE: These tracks have not been liftOver'd to original alts... at least not completely.
# They were lifted over from hg19, and some happened to be lifted from hg19 main chroms
# to hg38 initial assembly alts. While it would be nice to lift from hg38 main chroms
# to all alts and patches, that would lose the information that was mapped to some of the
# alts in the hg19-to-hg38 lift. So don't use mainToAllAltPatch, use mainToPatch and
# accept that the original alts will have a few good mappings from hg19, but mostly missing data.
# 18-08-10 with original files:
for dir in /hive/data/genomes/hg38/bed/hg19MassiveLift/wgEncodeReg/{*Mark*,*Txn}; do
composite=$(basename $dir)
echo $composite
cd $dir
for f in wg*.bigWig; do
track=$(basename $f .bigWig)
~/kent/src/hg/utils/liftOverBigWigToPatches $f \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
/hive/data/genomes/hg38/chrom.sizes \
$track.plusP11.bigWig &
done
wait
done
# 18-08-10: Install and rename original files
for dir in /hive/data/genomes/hg38/bed/hg19MassiveLift/wgEncodeReg/{*Mark*,*Txn}; do
composite=$(basename $dir)
echo $composite
cd $dir
for f in wg*.plusP11.bigWig; do
track=$(basename $f .plusP11.bigWig)
# First time only -- don't mv this in subsequent patches!
mv $track.bigWig $track.initial.bigWig
ln -sf `pwd`/$track.plusP11.bigWig /gbdb/hg38/bbi/wgEncodeReg/$composite/$track.bigWig
done
done
# 18-11-06, 19-01-07: Recompute .plusP11 files from .initial files with updated .chain file
for dir in /hive/data/genomes/hg38/bed/hg19MassiveLift/wgEncodeReg/{*Mark*,*Txn}; do
composite=$(basename $dir)
echo $composite
cd $dir
for f in wg*.initial.bigWig; do
track=$(basename $f .initial.bigWig)
~/kent/src/hg/utils/liftOverBigWigToPatches $f \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
/hive/data/genomes/hg38/chrom.sizes \
$track.plusP11.bigWig &
done
wait
done
# Not reinstalled -- used to recompute plusP12
##############################################################################
# Extend wgEncodeRegDnaseClustered (DONE 19-01-08 angie)
# first done 18-08-10; redone 18-11-06 and 19-01-08 with updated .over.chain file.
#NOTE: this has not been liftOver'd to original alts, aside from hg19 chr to -> hg38 alt!
cd /hive/data/genomes/hg38/bed/wgEncodeRegDnase
origFile=clusters/uwEnc2DnaseClustered.bed
liftOver -multiple -bedPlus=5 -noSerial $origFile \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
wgEncodeRegDnaseClustered.p11.bed /dev/null
sort -k1,1 -k2n,2n $origFile wgEncodeRegDnaseClustered.p11.bed \
> wgEncodeRegDnaseClustered.plusP11.bed
hgLoadBed -type=bed5+ -sqlTable=$HOME/kent/src/hg/lib/bed5SourceVals.sql -renameSqlTable \
hg38 wgEncodeRegDnaseClustered \
wgEncodeRegDnaseClustered.plusP11.bed
# 18-11-06, 19-01-08: Recompute .plusP11 files with updated .chain file.
cd /hive/data/genomes/hg38/bed/wgEncodeRegDnase
origFile=clusters/uwEnc2DnaseClustered.bed
liftOver -multiple -bedPlus=5 -noSerial $origFile \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
wgEncodeRegDnaseClustered.p11.bed /dev/null
sort -k1,1 -k2n,2n $origFile wgEncodeRegDnaseClustered.p11.bed \
> wgEncodeRegDnaseClustered.plusP11.bed
# Not reinstalled -- used to recompute plusP12
##############################################################################
# Extend wgEncodeRegTfbsClusteredV3 (DONE 19-01-08 angie)
# first done 18-08-10; redone 18-11-06 and 19-01-08 with updated .over.chain file.
#NOTE: this has not been liftOver'd to original alts, aside from hg19 chr to -> hg38 alt!
# 18-08-10 with original files:
cd /hive/data/genomes/hg38/bed/hg19MassiveLift/wgEncodeReg/wgEncodeRegTfbsClusteredV3/
origFile=hg38.wgEncodeRegClusteredV3.bed
liftOver -multiple -bedPlus=5 -noSerial $origFile \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
wgEncodeRegTfbsClusteredV3.p11.bed /dev/null
sort -k1,1 -k2n,2n $origFile wgEncodeRegTfbsClusteredV3.p11.bed \
> wgEncodeRegTfbsClusteredV3.plusP11.bed
hgLoadBed -type=bed5+ -sqlTable=$HOME/kent/src/hg/lib/bed5SourceVals.sql -renameSqlTable \
hg38 wgEncodeRegTfbsClusteredV3 wgEncodeRegTfbsClusteredV3.plusP11.bed
# 18-11-06, 19-01-08: Recompute .plusP11 files with updated .chain file.
cd /hive/data/genomes/hg38/bed/hg19MassiveLift/wgEncodeReg/wgEncodeRegTfbsClusteredV3/
origFile=hg38.wgEncodeRegClusteredV3.bed
liftOver -multiple -bedPlus=5 -noSerial $origFile \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
wgEncodeRegTfbsClusteredV3.p11.bed /dev/null
sort -k1,1 -k2n,2n $origFile wgEncodeRegTfbsClusteredV3.p11.bed \
> wgEncodeRegTfbsClusteredV3.plusP11.bed
# Not reinstalled -- used to recompute plusP12
##############################################################################
# Extend GTEX GENE (DONE 19-01-15 angie)
# first done 18-08-10; redone 18-11-06 and 19-01-08 with updated .over.chain file.
# gtexGeneModel redone 19-01-15 after implementing liftOver -multiple -genePred.
#NOTE: this has not been liftOver'd to original alts, aside from hg19 chr to -> hg38 alt!
# I'm not really sure what file(s) are the true source of the latest hg38 GTEX Gene tables,
# so I'll just work from the tables.
mkdir /hive/data/genomes/hg38/bed/gtex.p11
cd /hive/data/genomes/hg38/bed/gtex.p11
# There is actually no bin column in gtexGene.
hgsql hg38 -NBe 'select * from gtexGene' > gtexGene.initial.bed
liftOver -multiple -bedPlus=6 -noSerial gtexGene.initial.bed \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
gtexGene.p11.bed /dev/null
sort -k1,1 -k2n,2n gtexGene.initial.bed gtexGene.p11.bed \
| hgLoadBed -noBin -type=bed6+ -sqlTable=$HOME/kent/src/hg/lib/gtexGeneBed.sql -renameSqlTable \
hg38 gtexGene stdin
# gtexGeneModel does have a bin.
hgsql hg38 -NBe 'select * from gtexGeneModel' | cut -f 2- > gtexGeneModel.initial.gp
liftOver -multiple -genePred gtexGeneModel.initial.gp \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
gtexGeneModel.p11.gp /dev/null
sort -k2,2 -k3n,3n gtexGeneModel.initial.gp gtexGeneModel.p11.gp \
| hgLoadGenePred hg38 gtexGeneModel stdin
# 18-11-06, 19-01-08: Recompute .plusP11 files with updated .chain file.
# Don't reload tables because we're about to recompute plusP12.
cd /hive/data/genomes/hg38/bed/gtex.p11
liftOver -multiple -bedPlus=6 -noSerial gtexGene.initial.bed \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
gtexGene.p11.bed /dev/null
sort -k1,1 -k2n,2n gtexGene.initial.bed gtexGene.p11.bed \
> gtexGene.plusP11.bed
# 19-01-15: recompute now that liftOver -multiple -genePred works.
liftOver -multiple -genePred gtexGeneModel.initial.gp \
/hive/data/genomes/hg38/jkStuff/hg38.mainToPatch.p11.over.chain \
gtexGeneModel.p11.gp /dev/null
sort -k2,2 -k3n,3n gtexGeneModel.initial.gp gtexGeneModel.p11.gp \
> gtexGeneModel.plusP11.gp
##############################################################################
# Extend wgEncodeRegDnase (DNase HS) (DONE 19-01-23 angie)
# Nothing on the original alts, so use mainToAllAltPatch
# 95 Peak view subtracks
# I tried in vain to find the files from which the *Peak tables were loaded.
# The closest I got were some files like
# /hive/data/genomes/hg38/bed/wgEncodeRegDnase/run_pooled_hotspot5.1/wgEncodeEH000484.pooled.narrowPeak
# a bigBed file which has the same number of rows as wgEncodeRegDnaseUwK562Peak,
# and similar but not identical values.
# So, first dump values from the database tables.
mkdir /hive/data/genomes/hg38/bed/wgEncodeRegDnase/wgEncodeRegDnaseHS.p11
cd /hive/data/genomes/hg38/bed/wgEncodeRegDnase/wgEncodeRegDnaseHS.p11
for table in $(hgsql hg38 -NBe 'show tables like "wgEncodeRegDnase%Peak";'); do
echo $table
hgsql hg38 -NBe 'select * from '$table';' | cut -f 2- > $table.initial.bed
done
for f in *.initial.bed; do
track=$(basename $f .initial.bed)
echo $track
liftOver -multiple -bedPlus=5 -noSerial $f \
/hive/data/genomes/hg38/jkStuff/hg38.mainToAllAltPatch.p11.over.chain \
$track.p11.bed /dev/null
sort -k1,1 -k2n,2n $f $track.p11.bed > $track.plusP11.bed
done
# 95 Hotspots view subtracks
mkdir /hive/data/genomes/hg38/bed/wgEncodeRegDnase/wgEncodeRegDnaseHotspot.p11
cd /hive/data/genomes/hg38/bed/wgEncodeRegDnase/wgEncodeRegDnaseHotspot.p11
# Similar to wgEncodeRegDnaseWig, the original /gbdb/ links point to files with names
# that differ from the track names. Save that association and use it to add patches.
ls -l /gbdb/hg38/bbi/wgEncodeRegDnase/wgEncodeRegDnase*Hotspot*.bb \
| awk '{print $9 "\t" $11;}' > wgEncodeRegDnaseHotspot.gbdbToOrigFile.tab
wc -l wgEncodeRegDnaseHotspot.gbdbToOrigFile.tab
#95 wgEncodeRegDnaseHotspot.gbdbToOrigFile.tab
cat >runOne <<'_EOF_'
#!/bin/bash
set -beEu -o pipefail
track=$1
origFile=$2
bigBedToBed $origFile stdout \
| liftOver -multiple -bedPlus=6 -noSerial stdin \
/hive/data/genomes/hg38/jkStuff/hg38.mainToAllAltPatch.p11.over.chain \
$track.broadPeak.p11.bed /dev/null
sort -k1,1 -k2n,2n <(bigBedToBed $origFile stdout) $track.broadPeak.p11.bed \
> $track.broadPeak.plusP11.bed
# Don't make bigBed yet; we'll use the .bed to compute .plusP12 next, and install that.
#bedToBigBed -as $HOME/kent/src/hg/lib/bigNarrowPeak.as -type=bed6+3 $track.broadPeak.plusP11.bed \
# /hive/data/genomes/hg38/chrom.sizes $track.broadPeak.plusP11.bb
_EOF_
chmod a+x runOne
while read gbdbPath origFile; do
track=$(basename $gbdbPath .broadPeak.bb)
echo ./runOne $track $origFile
done < wgEncodeRegDnaseHotspot.gbdbToOrigFile.tab > jobList
para make jobList
para time
#Completed: 95 of 95 jobs
#CPU time in finished jobs: 153s 2.56m 0.04h 0.00d 0.000 y
#IO & Wait Time: 245s 4.08m 0.07h 0.00d 0.000 y
#Average job time: 4s 0.07m 0.00h 0.00d
#Longest finished job: 8s 0.13m 0.00h 0.00d
#Submission to last job: 26s 0.43m 0.01h 0.00d
# Don't do wgEncodeRegDnaseSignal view... the data files are same as DnaseWig below!
# Don't update /gbdb -- use .plusP11 files to make .plusP12.
##############################################################################
# Extend wgEncodeRegDnaseWig (DNase Signal) (DONE 19-01-23 angie)
# Nothing on the original alts, so use mainToAllAltPatch
mkdir /hive/data/genomes/hg38/bed/wgEncodeRegDnase/wgEncodeRegDnaseWig.p11
cd /hive/data/genomes/hg38/bed/wgEncodeRegDnase/wgEncodeRegDnaseWig.p11
# There are 95 subtracks. The original /gbdb/ links point to files with very different
# names from the track/link name. For example:
# /gbdb/hg38/bbi/wgEncodeRegDnase/wgEncodeRegDnaseUwHelas3Signal.bw ->
# /hive/data/genomes/hg38/bed/wgEncodeRegDnase/run_normalized/out/wgEncodeEH000495.norm.bw
# So for starters, make a file to record the correspondence between track and file.
ls -l /gbdb/hg38/bbi/wgEncodeRegDnase/wgEncodeRegDnase*Signal*.bw \
| awk '{print $9 "\t" $11;}' > wgEncodeRegDnaseWig.gbdbToOrigFile.tab
# Use that file to make a small cluster job to do the liftOvers
while read gbdbPath origFile; do
track=$(basename $gbdbPath .bw)
echo ~/kent/src/hg/utils/liftOverBigWigToPatches $origFile \
/hive/data/genomes/hg38/jkStuff/hg38.mainToAllAltPatch.p11.over.chain \
/hive/data/genomes/hg38/chrom.sizes \
{check out exists $track.plusP11.bw}
done < wgEncodeRegDnaseWig.gbdbToOrigFile.tab > jobList
para make jobList
para time
#Completed: 95 of 95 jobs
#CPU time in finished jobs: 136582s 2276.37m 37.94h 1.58d 0.004 y
#IO & Wait Time: 0s 0.00m 0.00h 0.00d 0.000 y
#Average job time: 463s 7.72m 0.13h 0.01d
#Longest finished job: 884s 14.73m 0.25h 0.01d
#Submission to last job: 1684s 28.07m 0.47h 0.02d
# Don't update /gbdb -- use .plusP11 files to make .plusP12.
+#############################################################################
+# Update hg38.p11.chromAlias.txt (DONE 2024-01-17 Angie)
+
+ # In MLQ#32874, the user reported that chrUn_KI270752v1 is missing from hg38.p12.chromAlias.txt.
+ # That's because when the hg38.{p11,p12,p13}.chromAlias.txt files were initially created
+ # in patchUpdate.13.txt ("Correctly versioned hg38.chromAlias.txt files in downloads"),
+ # chrUn_KI270752v1 was omitted because it had been removed from the RefSeq assembly as
+ # contamination. However, it is confusing for users to have a sequence in the db with no
+ # aliases despite it having Assembly, Ensembl and INSDC aliases. So add it back.
+ hgsql hg38 -NBe 'select * from chromAlias where chrom = "chrUn_KI270752v1"'
+#HSCHRUN_RANDOM_CTG29 chrUn_KI270752v1 assembly
+#KI270752.1 chrUn_KI270752v1 ensembl,genbank
+ cd /hive/data/genomes/hg38/goldenPath/bigZips
+ echo -e "chrUn_KI270752v1\tHSCHRUN_RANDOM_CTG29\tKI270752.1\t" >> hg38.p11.chromAlias.txt
+ # p11/hg38.p11.chromAlias.txt is a symlink to the one in this directory.
+ # p11/hg38.p11.chromAlias.bb is a file in p11/ . -- But it already has chrUn_KI270752v1
+ # so it does not need to be updated, great.
+ bigBedToBed -chrom=chrUn_KI270752v1 p11/hg38.p11.chromAlias.bb stdout
+#chrUn_KI270752v1 0 27745 chrUn_KI270752v1 HSCHRUN_RANDOM_CTG29 KI270752.1 KI270752.1
+
+
##############################################################################