a27219ced37a549f3b69bc79f07ea99a87439ecd
braney
  Sat Feb 20 12:49:06 2021 -0800
Give user a chance to get non-short circuit hits if there's more than
one short circuit choice (otherwise just go to position).   Add ability
to add trixSearch path to findSpecs.   Add new find spec for hg38
gencodeV36 that only has canonical genes in it.

diff --git src/hg/makeDb/doc/ucscGenes/hg38.gencodeV36.sh src/hg/makeDb/doc/ucscGenes/hg38.gencodeV36.sh
index 10e5c32..dfd333a 100644
--- src/hg/makeDb/doc/ucscGenes/hg38.gencodeV36.sh
+++ src/hg/makeDb/doc/ucscGenes/hg38.gencodeV36.sh
@@ -349,30 +349,36 @@
     hgLoadRnaFold -warnEmpty $tempDb foldUtr3 fold
 
 # Clean up
     rm -r split fold err batch.bak
     cd ../utr5
     rm -r split fold err batch.bak
 
 hgKgGetText $tempDb tempSearch.txt
 sort tempSearch.txt > tempSearch2.txt
 tawk '{split($2,a,"."); printf "%s\t", $1;for(ii = 1; ii <= a[2]; ii++) printf "%s ",a[1] "." ii; printf "\n" }' txToAcc.tab | sort > tempSearch3.txt
 join tempSearch2.txt tempSearch3.txt | sort > knownGene.txt
 ixIxx knownGene.txt knownGene${GENCODE_VERSION}.ix knownGene${GENCODE_VERSION}.ixx
  rm -rf /gbdb/$db/knownGene${GENCODE_VERSION}.ix /gbdb/$db/knownGene${GENCODE_VERSION}.ixx
 ln -s $dir/knownGene${GENCODE_VERSION}.ix  /gbdb/$db/knownGene${GENCODE_VERSION}.ix
 ln -s $dir/knownGene${GENCODE_VERSION}.ixx /gbdb/$db/knownGene${GENCODE_VERSION}.ixx  
+tawk '{print $5}' knownCanonical.tab | sort > knownCanonicalId.txt
+join knownCanonicalId.txt knownGene.txt > knownGeneFast.txt
+ixIxx knownGeneFast.txt knownGeneFast${GENCODE_VERSION}.ix knownGeneFast${GENCODE_VERSION}.ixx
+ rm -rf /gbdb/$db/knownGeneFast${GENCODE_VERSION}.ix /gbdb/$db/knownGeneFast${GENCODE_VERSION}.ixx
+ln -s $dir/knownGeneFast${GENCODE_VERSION}.ix  /gbdb/$db/knownGeneFast${GENCODE_VERSION}.ix
+ln -s $dir/knownGeneFast${GENCODE_VERSION}.ixx /gbdb/$db/knownGeneFast${GENCODE_VERSION}.ixx  
 
 #zcat gencode${GENCODE_VERSION}.bed.gz > ucscGenes.bed
 #jtwoBitToFa -noMask /cluster/data/$db/$db.2bit -bed=ucscGenes.bed stdout | faFilter -uniq stdin  ucscGenes.fa
 #jhgPepPred $tempDb generic knownGeneMrna ucscGenes.fa
 bedToPsl /cluster/data/$db/chrom.sizes ucscGenes.bed ucscGenes.psl
 pslRecalcMatch ucscGenes.psl /cluster/data/$db/$db.2bit ucscGenes.fa kgTargetAli.psl
 # should be zero
 awk '$11 != $1 + $3+$4' kgTargetAli.psl
 hgLoadPsl $tempDb kgTargetAli.psl
 
 cd $dir
 # Make PCR target for UCSC Genes, Part 1.
 # 1. Get a set of IDs that consist of the UCSC Gene accession concatenated with the
 #    gene symbol, e.g. uc010nxr.1__DDX11L1
 hgsql $tempDb -N -e 'select kgId,geneSymbol from kgXref' \