d6677c2b706a2424e70bcacd517d8a61f4b39f20 angie Wed Apr 21 15:09:20 2021 -0700 Apparently GISAID sequence names can have apostrophes too. Strip out more troublemaking characters. diff --git src/hg/utils/otto/sarscov2phylo/gisaidFromChunks.sh src/hg/utils/otto/sarscov2phylo/gisaidFromChunks.sh index 58060e4..1bd7e1a 100755 --- src/hg/utils/otto/sarscov2phylo/gisaidFromChunks.sh +++ src/hg/utils/otto/sarscov2phylo/gisaidFromChunks.sh @@ -6,34 +6,35 @@ # kent/src/hg/utils/otto/sarscov2phylo/gisaidFromChunks.sh # Make nextfasta and nextmeta substitute files from chunks of downloaded GISAID sequences lastRealNextmeta=metadata_2020-12-08_20-35.tsv.gz today=$(date +%F) # Run pangolin and nextclade on any chunks that need it cd /hive/users/angie/gisaid/chunks make cd /hive/users/angie/gisaid # Glom all the chunks together. # Remove initial "hCoV-19/" and remove spaces a la nextmeta (e.g. "Hong Kong" -> "HongKong"). +# Strip single quotes (e.g. "Cote d'Ivoire" --> "CotedIvoire"). # Also remove a stray comma in a name that caused Newick parsing error ("Hungary/US-32533w,/2020"). # Keep the strain|epiId|date "full names". time xzcat chunks/gisaid_epi_isl_*.fa.xz \ -| sed -re 's@^>hCo[Vv]-19/+@>@; s/ //g; s/,//; s/\r$//;' \ +| sed -re 's@^>hCo[Vv]-19/+@>@; s/[ '"'"',()]//g; s/\r$//;' \ | xz -T 50 \ > gisaid_fullNames_$today.fa.xz # Make tmp files with a fullName key and various columns that we'll join together. fastaNames gisaid_fullNames_$today.fa.xz \ | awk -F\| -vOFS="\t" '{print $0, $1, $2, $3;}' \ | sort \ > tmp.first3 # Sequence length faSize -detailed <(xzcat gisaid_fullNames_$today.fa.xz) | sort > tmp.lengths # Lineage & clade assignments sort chunks/pangolin.tsv \ > tmp.lineage sort chunks/nextclade.tsv \ > tmp.clade