introduce MIN_HMM_EVALUE constant

oschwengers · Sep 24, 2024 · 4f40f46 · 4f40f46
1 parent 870a427
commit 4f40f46
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 1 deletion.
diff --git a/bakta/constants.py b/bakta/constants.py
@@ -26,6 +26,7 @@
 MIN_PSC_IDENTITY = 0.9  # min protein identity for PSC detection
 MIN_SORF_COVERAGE = 0.9  # min sORF coverage for PSC detection
 MIN_SORF_IDENTITY = 0.9  # min sORF identity for PSC detection
+MIN_HMM_EVALUE = 1e-6  # min evalue for CDS HMM searches
 HYPOTHETICAL_PROTEIN = 'hypothetical protein'  # hypothetical protein product description
 CDS_MAX_OVERLAPS = 30  # max overlap [bp] allowed for user-provided/de novo-predicted CDS overlaps
 CDS_SOURCE_USER = 'user-provided'

diff --git a/bakta/features/orf.py b/bakta/features/orf.py
@@ -24,7 +24,8 @@ def detect_spurious(orfs: Sequence[dict]):
         for top_hits in pyhmmer.hmmsearch(hmm, proteins, bit_cutoffs='gathering', cpus=cfg.threads):
             for hit in top_hits:
                 orf = orf_by_aa_digest[hit.name.decode()]
-                if hit.evalue > 1E-5:
+                print(f'ANTIFAM: hit.evalue={hit.evalue}')
+                if hit.evalue > bc.MIN_HMM_EVALUE:
                     log.debug(
                         'discard low spurious E value: contig=%s, start=%i, stop=%i, strand=%s, subject=%s, evalue=%1.1e, bitscore=%f',
                         orf['contig'], orf['start'], orf['stop'], orf['strand'], hit.best_domain.alignment.hmm_name.decode(), hit.evalue, hit.score