small pipeline change to match prepare-blast behav

Author: dsmutin

README.md

@@ -31,11 +31,11 @@ PROBEst can be run using the following command:
 
 ```bash
 python pipeline.py \
-  -i {INPUT} \
-  -tb {TRUE_BASE} \
-  -fb [FALSE_BASE ...] \
-  -c {CONTIG_TABLE} \
-  -o {OUTPUT}
+  -i {INPUT} \ # fasta // .fa.gz
+  -tb {TRUE_BASE} \ # directory with blastn database / fasta files
+  -fb [FALSE_BASE ...] \ # directories with blastn database / fasta files
+  -c {CONTIG_TABLE} \ # .tsv table with BLAST database information (optional; defaults to `{OUTPUT}/contigs.tsv` when using FASTA directories).
+  -o {OUTPUT} \ # output directory
 ```
 
 **Blastn databases** and **contig table** are produced by ```prep_db.sh``` (or built automatically when `-tb` / `-fb` point at **directories of FASTA files**). After each automatic build, the contig table is **deduplicated by contig ID** (duplicate rows from merged FASTAs or repeated runs keep the last mapping).

pipeline.py

@@ -182,10 +182,11 @@ def merge_iter(iter: int):
 
     print("Positive hits counted")
 
-    # false bases
-    for db_neg in args.false_base:
-        blastn_db = blastn_iter + " -db " + db_neg + \
-            " >> " + out_dir(iter) + "negative_hits.tsv"
+    # false bases: first DB uses '>' so negative_hits.tsv is not appended to stale files
+    neg_path = out_dir(iter) + "negative_hits.tsv"
+    for i, db_neg in enumerate(args.false_base):
+        redir = " > " if i == 0 else " >> "
+        blastn_db = blastn_iter + " -db " + db_neg + redir + neg_path
         subprocess.run(blastn_db, shell=True)
 
     print("Negative hits counted")

src/PROBESt/args.py

@@ -73,7 +73,7 @@ def arguments_parse():
     parser.add_argument("-c", "--contig_table",
                         required=False,
                         default=None,
-                        help="Path to a .tsv table containing BLAST database information. If not provided and FASTA directories are used for bases, it will be auto-generated in the output directory.")
+                        help="Path to the genome/contig .tsv (prep_db.sh -c format). If omitted, defaults to <output>/contigs.tsv. When -tb/-fb are FASTA directories, all bases contribute rows to this file (same as running prep_db.sh with one shared -c).")
 
     parser.add_argument("-o", "--output",
                         required=True,

tests/test_pipeline_general_integration.py

@@ -23,7 +23,7 @@
 # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 # SOFTWARE.
 
-"""Integration test equivalent to repo root test.sh (general FISH dataset)."""
+"""Integration test: general FISH dataset with JIT FASTA bases, AI + visualization."""
 
 import shutil
 import subprocess
@@ -42,8 +42,8 @@ def _require_blast_tools():
 
 
 @pytest.mark.integration
-def test_pipeline_general_fish_from_test_sh(tmp_path):
-    """Mirror `test.sh` snippet: small general run (AI enabled by default, visualize disabled)."""
+def test_integration_fish(tmp_path):
+    """Same CLI as: …/test.fna, fasta_base true/false dirs, -N 3, --visualize True --AI True."""
     _require_blast_tools()
 
     output = tmp_path / "output"
@@ -67,6 +67,10 @@ def test_pipeline_general_fish_from_test_sh(tmp_path):
         "5",
         "-N",
         "3",
+        "--visualize",
+        "True",
+        "--AI",
+        "True",
     ]
 
     result = subprocess.run(
@@ -83,5 +87,6 @@ def test_pipeline_general_fish_from_test_sh(tmp_path):
 
     assert (output / "modeling_results.tsv").is_file()
     assert (output / "output_dedegenerated.fa").is_file()
-    # Visualizations are only created when `--visualize` is enabled.
-    assert not (output / "visualizations").exists()
+    viz_dir = output / "visualizations"
+    assert viz_dir.is_dir()
+    assert any(viz_dir.glob("*_visualization.png"))