7ec003fc7ae8f5bb327037388ac9ce1031286b01
chmalee
  Thu Sep 17 12:32:59 2020 -0700
Making lovd long variants a bigBed 9 + so mergeSpannedItems works , refs #25841

diff --git src/hg/utils/otto/lovd/checkAndLoad.sh src/hg/utils/otto/lovd/checkAndLoad.sh
index 925ca02..5460b23 100755
--- src/hg/utils/otto/lovd/checkAndLoad.sh
+++ src/hg/utils/otto/lovd/checkAndLoad.sh
@@ -26,41 +26,42 @@
 fi
 if [ "$new19Cc" -ne "0" ]; then
         echo LVOD hg19 $today: found rows with not six columns, quitting
         exit 255
 fi
 
 for db in hg38 hg19
 do
     if [ ${db} == "hg19" ]
     then
         $KENTBIN/bedClip lovd.${db}.bed /cluster/data/${db}/chrom.sizes stdout | sed -e 's/^chrM/chrMT/g' > lovd.${db}.clipped.bed
     else
         $KENTBIN/bedClip lovd.${db}.bed /cluster/data/${db}/chrom.sizes lovd.${db}.clipped.bed
     fi
 
-    sort -k1,1 -k2,2n lovd.${db}.clipped.bed | awk '(($3-$2)<50)' > lovd.${db}.short.bed
-    sort -k1,1 -k2,2n lovd.${db}.clipped.bed | awk '(($3-$2)>=50)' > lovd.${db}.long.bed
+    sort -k1,1 -k2,2n lovd.${db}.clipped.bed | tawk '(($3-$2)<50)' > lovd.${db}.short.bed
+    # the long variants turn into bed9+ so that mergeSpannedItems works:
+    sort -k1,1 -k2,2n lovd.${db}.clipped.bed | tawk '(($3-$2)>=50){print $1,$2,$3,$4,0,".",$2,$3,"0,0,0",$5,$6,$7}' > lovd.${db}.long.bed
     if [ ${db} == "hg19" ]
     then
         echo -e 'chrM\t0\t16571\tCheck chrMT\tPlease look at chrMT, not chrM, for LOVD annotations.\t\t' >> lovd.${db}.short.bed
-        echo -e 'chrM\t0\t16571\tCheck chrMT\tPlease look at chrMT, not chrM, for LOVD annotations.\t\t' >> lovd.${db}.long.bed
+        echo -e 'chrM\t0\t16571\tCheck chrMT\t0\t.\t0\t16571\t0,0,0\tPlease look at chrMT, not chrM, for LOVD annotations.\t\t' >> lovd.${db}.long.bed
     fi
 
     sort -k1,1 -k2,2n lovd.${db}.short.bed > lovd.${db}.short.bed.sorted
     sort -k1,1 -k2,2n lovd.${db}.long.bed > lovd.${db}.long.bed.sorted
     oldShortLc=`bigBedToBed ../release/${db}/lovd.${db}.short.bb stdout | wc -l`
     oldLongLc=`bigBedToBed ../release/${db}/lovd.${db}.long.bb stdout | wc -l`
     newShortLc=`wc -l lovd.${db}.short.bed.sorted | cut -d' ' -f1`
     newLongLc=`wc -l lovd.${db}.long.bed.sorted | cut -d' ' -f1`
     echo ${db} short rowcount: old $oldShortLc new: $newShortLc
     echo ${db} long rowcount: old $oldLongLc new: $newLongLc
     echo $oldShortLc $newShortLc | awk -v d=${db} '{if (($2-$1)/$1 > 0.1) {printf "validate on %s LOVD short failed: old count: %d, new count: %d\n", d,$1,$2; exit 1;}}'
     echo $oldLongLc $newLongLc | awk -v d=${db} '{if (($2-$1)/$1 > 0.1) {printf "validate on %s LOVD long failed: old count: %d, new count: %d\n", d,$1,$2; exit 1;}}'
 
-    bedToBigBed -type=bed4+3 -tab -as=../lovd.as lovd.${db}.short.bed.sorted /cluster/data/${db}/chrom.sizes lovd.${db}.short.bb
-    bedToBigBed -type=bed4+3 -tab -as=../lovd.as lovd.${db}.long.bed.sorted /cluster/data/${db}/chrom.sizes lovd.${db}.long.bb
+    bedToBigBed -type=bed4+3 -tab -as=../lovd.short.as lovd.${db}.short.bed.sorted /cluster/data/${db}/chrom.sizes lovd.${db}.short.bb
+    bedToBigBed -type=bed9+3 -tab -as=../lovd.long.as lovd.${db}.long.bed.sorted /cluster/data/${db}/chrom.sizes lovd.${db}.long.bb
     mkdir -p ${WORKDIR}/release/${db}
     cp lovd.${db}.{short,long}.bb ${WORKDIR}/release/${db}/
 done
 
 echo LOVD update: OK