Fix ignoring intact sequences in table_precursor and landscapes_plots

cfe-lab · Jul 29, 2024 · 87aaffb · 87aaffb
1 parent 9454ade
commit 87aaffb
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 5 deletions.
diff --git a/gene_splicer/landscapes.py b/gene_splicer/landscapes.py
@@ -56,14 +56,15 @@ def generate_proviral_landscape_csv_1_cont(blastn_reader: csv.DictReader,
             ref_end = new_end
             is_inverted = 'yes'
 
-        verdict = verdicts.get(qseqid)
+        verdict = verdicts[qseqid]
+        is_defective = verdict != 'Intact'
         landscape_entry = {'ref_start': ref_start,
                            'ref_end': ref_end,
                            'samp_name': sample_name,
                            'run_name': run_name,
                            'is_inverted': is_inverted,
-                           'is_defective': verdict is not None,
-                           'defect': verdict or 'Intact',
+                           'is_defective': is_defective,
+                           'defect': verdict,
                            }
 
         landscape_writer.writerow(landscape_entry)

diff --git a/gene_splicer/utils.py b/gene_splicer/utils.py
@@ -428,15 +428,21 @@ def iterate_hivintact_verdicts_1(directory: Path, intact: Set[str] = set()) -> I
     intact = set()
 
     def get_verdict(SEQID: str, all_errors) -> Tuple[str, str]:
-        ordered = sorted(all_errors, key=HIVINTACT_ERRORS_TABLE.index)
-        verdict = ordered[0]
+        if all_errors:
+            ordered = sorted(all_errors, key=HIVINTACT_ERRORS_TABLE.index)
+            verdict = ordered[0]
+        else:
+            verdict = "Intact"
+
         return (SEQID, verdict)
 
     with open(os.path.join(directory, 'holistic.csv'), 'r') as f:
         reader = csv.DictReader(f)
         for row in reader:
             if row["intact"] == "True":
                 intact.add(row["qseqid"])
+                SEQID = row["qseqid"]
+                yield get_verdict(SEQID, all_errors=[])
 
     with open(os.path.join(directory, 'defects.csv'), 'r') as f:
         reader = csv.DictReader(f)