72739f248f43d3685eb2ca5fd464f73102fb77a4
angie
  Thu Jan 13 11:36:29 2022 -0800
Increase split fasta size for nextclade run - more efficient.

diff --git src/hg/utils/otto/sarscov2phylo/getNcbi.sh src/hg/utils/otto/sarscov2phylo/getNcbi.sh
index 17b617b..79e9964 100755
--- src/hg/utils/otto/sarscov2phylo/getNcbi.sh
+++ src/hg/utils/otto/sarscov2phylo/getNcbi.sh
@@ -72,31 +72,31 @@
 
 # Run pangolin and nextclade on sequences that are new since yesterday
 export TMPDIR=/dev/shm
 fastaNames genbank.fa.xz | awk '{print $1;}' | sed -re 's/\|.*//' | grep -vx pdb | sort -u > gb.names
 splitDir=splitForNextclade
 rm -rf $splitDir
 mkdir $splitDir
 if [ -e ../ncbi.latest/nextclade.tsv ]; then
     cp ../ncbi.latest/nextclade.tsv .
     cut -f 1 nextclade.tsv | sort -u > nextclade.prev.names
     comm -23 gb.names nextclade.prev.names > nextclade.names
     faSomeRecords <(xzcat genbank.fa.xz) nextclade.names nextclade.fa
     faSplit about nextclade.fa 30000000 $splitDir/chunk
 else
     cp /dev/null nextclade.tsv
-    faSplit about <(xzcat genbank.fa.xz) 30000000 $splitDir/chunk
+    faSplit about <(xzcat genbank.fa.xz) 300000000 $splitDir/chunk
 fi
 if (( $(ls -1 splitForNextclade | wc -l) > 0 )); then
     nDataDir=~angie/github/nextclade/data/sars-cov-2
     outDir=$(mktemp -d)
     outTsv=$(mktemp)
     for chunkFa in $splitDir/chunk*.fa; do
         nextclade -j 50 -i $chunkFa \
             --input-root-seq $nDataDir/reference.fasta \
             --input-tree $nDataDir/tree.json \
             --input-qc-config $nDataDir/qc.json \
             --output-dir $outDir \
             --output-tsv $outTsv >& nextclade.log
         cut -f 1,2 $outTsv | tail -n+2 | sed -re 's/"//g;' >> nextclade.tsv
         rm $outTsv
     done