9fb2b35720b02cca3a9cece5cc8cdb782f86ed34
angie
  Tue Oct 31 09:56:35 2023 -0700
Compress some previously uncompressed files so we don't waste quite so much disk space.  Also spare sequences listed in sars-cov-2-variants/lineage-proposals repo file recombinants.tsv from the nextclade-based filters.

diff --git src/hg/utils/otto/sarscov2phylo/getNcbi.sh src/hg/utils/otto/sarscov2phylo/getNcbi.sh
index f96a26b..57fba4d 100755
--- src/hg/utils/otto/sarscov2phylo/getNcbi.sh
+++ src/hg/utils/otto/sarscov2phylo/getNcbi.sh
@@ -125,39 +125,39 @@
     fa=$1
     out=$fa.pangolin.csv
     logfile=$(mktemp)
     pangolin -t 4 $fa --skip-scorpio --outfile $out > $logfile 2>&1
     rm $logfile
 }
 export -f runPangolin
 if [ -e ../ncbi.latest/lineage_report.csv ]; then
     cp ../ncbi.latest/lineage_report.csv linRepYesterday
     tail -n+2 linRepYesterday | sed -re 's/^([A-Z]+[0-9]+\.[0-9]+).*/\1/' | sort \
         > pangolin.prev.names
     comm -23 gb.names pangolin.prev.names > pangolin.names
     faSomeRecords <(xzcat genbank.fa.xz) pangolin.names stdout \
     | sed -re '/^>/  s/^>([A-Z]{2}[0-9]{6}\.[0-9]+) \|[^,]+/>\1/;' \
         > pangolin.fa
-    pangolin --analysis-mode pangolearn pangolin.fa >& pangolin.log
+    pangolin --skip-scorpio pangolin.fa >& pangolin.log
     tail -n+2 lineage_report.csv >> linRepYesterday
     mv linRepYesterday lineage_report.csv
     rm -f pangolin.fa
 else
     splitDir=splitForPangolin
     rm -rf $splitDir
     mkdir $splitDir
-    faSplit about <(xzcat genbank.fa.xz) 30000000 $splitDir/chunk
+    faSplit about <(xzcat genbank.fa.xz | sed -re '/^>/ s/ .*//;') 30000000 $splitDir/chunk
     find $splitDir -name chunk\*.fa \
     | parallel -j 10 "runPangolin {}"
     head -1 $(ls -1 $splitDir/chunk*.csv | head -1) > lineage_report.csv
     for f in $splitDir/chunk*.csv; do
         tail -n+2 $f >> lineage_report.csv
     done
     rm -rf $splitDir
 fi
 wc -l lineage_report.csv
 
 # It turns out that sometimes new sequences sneak into ncbi_dataset.tsv before they're included in
 # genomic.fna.  Filter those out so we don't have missing pangolin and nextclade for just-added
 # COG-UK sequences.  (https://github.com/theosanderson/taxodium/issues/10#issuecomment-876800176)
 grep -Fwf gb.names ncbi_dataset.plusBioSample.tsv > tmp
 wc -l tmp ncbi_dataset.plusBioSample.tsv