61309f7c6fb5fb1d582fde243fbfa83559c1da21
chmalee
  Tue Oct 5 16:26:13 2021 -0700
Fix float parsing bug in gtex transcript expression track on hg19 and hg38, refs #28280

diff --git src/hg/makeDb/doc/hg38/hg38.txt src/hg/makeDb/doc/hg38/hg38.txt
index 03a1488..8f065f8 100644
--- src/hg/makeDb/doc/hg38/hg38.txt
+++ src/hg/makeDb/doc/hg38/hg38.txt
@@ -6428,44 +6428,50 @@
 
 # Make a clean matrix
 cut -f 1 ../johnVivianRecompute/gtex.tpm.tab | cut -f 1 -d "." > gtexTranscripts.txt
 cut -f 2- ../johnVivianRecompute/gtex.tpm.tab > gtexTpmValues.tsv
 paste gtexTranscripts.txt gtexTpmValues.tsv > gtexMatrix.tsv
 rowsToCols gtexMatrix.tsv tspsdGtexMatrix.tsv
 sort tspsdGtexMatrix.tsv > sortedTspsdGtexMatrix.tsv
 grep -v '(CML)' gtexSamples.tsv | cut -f 1 | sed 's/Run_s/#transcript/g' | sort > sortedCleanGtexSamples.tsv
 join sortedCleanGtexSamples.tsv sortedTspsdGtexMatrix.tsv > cleanTspsdGtexMatrix.tsv
 rowsToCols cleanTspsdMatrix.tsv cleanGtexMatrix.tsv
 
 # Build a coordinate map
 hgsql hg38 -e "select * from ensGene" | cut -f 2- | sort > ensGene
 hgsql hg38 -e "select * from ensemblToGeneName" | sort >  ensemblToGeneName
 join ensGene ensemblToGeneName | awk '{print $2"\t"$4"\t"$5"\t"$1"\t0\t"$3"\t"$16}' > coord.bed
+# NOTE: CHRISL10-05-2021 - the above ensGene steps weren't actually done or the files were removed,
+# there was a coord.tsv which I used instead so the below re-run could work
+tawk '{print $1,$2,$3,$4,0,$5,$6}' coord.tsv > coord.bed
+# END CHRISL10-05-2021 NOTE)
 
 # Get the gtex ordering
 hgsql hgFixed -e "select * from gtexTissue" | cut -f 3 | sed 's/ - /-/g' | sed 's/ /_/g' | sed '1D' > gtexGroupOrder.txt
 
 # Use the meta data, matrix, and coordinate map to generate a barchart bed
+# NOTE: CHRISL10-05-2021 - re-ran this step to fix float parsing bug:
 time expMatrixToBarchartBed cleanGtexSamples.tsv cleanGtexMatrix.tsv coord.bed gtexTransExp.bed --groupOrderFile gtexGroupOrder.txt
 
 # NOTE: Use the header line of the bed file to populate the barChartBars field in the trackDb.
 # The order of the labels in the barChartBars field should match the order of the labels in the
 # expScores column in the bed file header.
 
 # Sort and convert into a bigBed file.
 sort -k1,1 -k2,2n gtexTransExp.bed > sortedGtexTransExp.bed
-bedToBigBed -as=$HOME/kent/src/hg/lib/barChartTranscExp.as -type=bed6+5 sortedGtexTransMed.bed /hive/data/genomes/hg38/chrom.sizes gtexTranscExpr.bb
+# NOTE: CHRISL10-05-2021 - re-ran bedToBigBed step with correct file names
+bedToBigBed -as=$HOME/kent/src/hg/lib/barChartBed.as -type=bed6+5 sortedGtexTransExp.bed /hive/data/genomes/hg38/chrom.sizes gtexTranscExpr.bb
 
 # Link the files into gbdb
 cd /gbdb/hgFixed/human/expMatrix
 ln -s /hive/data/outside/gtex/barChartTrack/cleanGtexSamples.tsv cleanGtexSamples.tab
 ln -s /hive/data/outside/gtex/barChartTrack/cleanGtexMatrix.tsv cleanGtexMatris.tab
 
 # <2007-08-30 kate)
 cd /gbdb/hg38/gtex
 ln -s /hive/data/outside/gtex/barChartTrack/gtexTranscExpr.bb .
 
 #########################################################################
 # LASTZ human/hg38 vs. Zebrafish /danRer11
 #	(DONE - 2017-06-12 - Chris)
 
     mkdir /hive/data/genomes/hg38/bed/lastzDanRer11.2017-06-12