feat: add configurable chunk size option to minos_config.yaml for bla…

…st/diamond jobs to reduce HPC overhead
EI-CoreBioinformatics · Feb 12, 2024 · 8fed890 · 8fed890
1 parent 4f6407d
commit 8fed890
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 1 deletion.
diff --git a/minos/etc/minos_config.yaml b/minos/etc/minos_config.yaml
@@ -28,13 +28,15 @@ params:
     quant: "-b 100"
   blast:
     makeblastdb: "-dbtype prot"
+    chunk_size: 1000
     blastp: "-max_target_seqs 1 -evalue 1e-5"
     blastx: "-max_target_seqs 1 -evalue 1e-5"
     tophit:
       pident_threshold: 0.0
       qcov_threshold: 0.0
   diamond:
     makedb: ""
+    chunk_size: 20000
     blastp: "--max-target-seqs 1 --evalue 1e-5"
     blastx: "--max-target-seqs 1 --evalue 1e-5"
     tophit:

diff --git a/minos/zzz/minos_run.smk b/minos/zzz/minos_run.smk
@@ -473,7 +473,7 @@ checkpoint minos_chunk_proteins:
 	log:
 		os.path.join(LOG_DIR, os.path.basename(rules.minos_gffread_extract_sequences.output.pep if config["blast-mode"] == "blastp" else rules.minos_gffread_extract_sequences.output.cds) + ".chunk.log")
 	params:
-		chunksize = 1000,
+		chunksize = config["params"]["diamond"]["chunk_size"] if config["use-diamond"] else config["params"]["blast"]["chunk_size"],
 		outdir = os.path.join(TEMP_DIR, "chunked_proteins")
 	threads:
 		HPC_CONFIG.get_cores("minos_chunk_proteins")