Skip to content

Commit

Permalink
Fix database creation for GTDB r214 (#742)
Browse files Browse the repository at this point in the history
* Fix database creation for GTDB r214

* Remove the .gz extension when renaming entries in the lookup file

* Fix VERSION.txt file name for GTDB r214
  • Loading branch information
apcamargo authored Aug 25, 2023
1 parent ad6dfc6 commit 4b52296
Showing 1 changed file with 3 additions and 3 deletions.
6 changes: 3 additions & 3 deletions data/workflow/databases.sh
Original file line number Diff line number Diff line change
Expand Up @@ -136,7 +136,7 @@ case "${SELECTION}" in
;;
"GTDB")
if notExists "${TMP_PATH}/download.done"; then
downloadFile "https://data.ace.uq.edu.au/public/gtdb/data/releases/latest/VERSION" "${TMP_PATH}/version"
downloadFile "https://data.ace.uq.edu.au/public/gtdb/data/releases/latest/VERSION.txt" "${TMP_PATH}/version"
downloadFile "https://data.ace.uq.edu.au/public/gtdb/data/releases/latest/genomic_files_reps/gtdb_proteins_aa_reps.tar.gz" "${TMP_PATH}/gtdb.tar.gz"
downloadFile "https://data.ace.uq.edu.au/public/gtdb/data/releases/latest/bac120_taxonomy.tsv" "${TMP_PATH}/bac120_taxonomy.tsv"
downloadFile "https://data.ace.uq.edu.au/public/gtdb/data/releases/latest/ar53_taxonomy.tsv" "${TMP_PATH}/ar53_taxonomy.tsv"
Expand Down Expand Up @@ -371,9 +371,9 @@ case "${INPUT_TYPE}" in
;;
"GTDB")
# shellcheck disable=SC2086
"${MMSEQS}" tar2db "${TMP_PATH}/gtdb.tar.gz" "${TMP_PATH}/tardb" --tar-include 'faa$' ${THREADS_PAR} \
"${MMSEQS}" tar2db "${TMP_PATH}/gtdb.tar.gz" "${TMP_PATH}/tardb" --tar-include 'faa.gz$' ${THREADS_PAR} \
|| fail "tar2db died"
sed 's|_protein\.faa||g' "${TMP_PATH}/tardb.lookup" > "${TMP_PATH}/tardb.lookup.tmp"
sed 's|_protein\.faa\.gz||g' "${TMP_PATH}/tardb.lookup" > "${TMP_PATH}/tardb.lookup.tmp"
mv -f -- "${TMP_PATH}/tardb.lookup.tmp" "${TMP_PATH}/tardb.lookup"
# shellcheck disable=SC2086
"${MMSEQS}" createdb "${TMP_PATH}/tardb" "${OUTDB}" ${COMP_PAR} \
Expand Down

0 comments on commit 4b52296

Please sign in to comment.