Harden cache_msa: pick valid FASTA a3m + strip null bytes

Pilot validated the MSA-reuse architecture (1 server query, cached &
reused -- server hammering fixed) but Boltz's INPUT a3m parser crashed on
the cached file: KeyError '\x00'. Boltz writes several .a3m files (some
binary); we were grabbing a bad one.

Fix: select the largest a3m that looks like FASTA (starts with '>'), and
strip null bytes before caching. Raise with the candidate list if none
qualifies (so we can see what's there).

Pilot status: MSA caching + reuse confirmed working; this fixes the
a3m-format crash. Re-validate with a 2-drug pilot before the full run.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

gpu/modal_app.py

@@ -148,14 +148,22 @@ def cache_msa(label: str, protein_seq: str, ligand_smiles: str, cofactor_ccds: l
     out = work / "out"
     subprocess.run(["boltz", "predict", str(work / "in.yaml"), "--use_msa_server",
                     "--cache", boltz_cache, "--out_dir", str(out), "--output_format", "pdb"], check=True)
-    a3m = next(out.rglob("*.a3m"), None)
+    # Pick the largest a3m that actually looks like FASTA/a3m text (Boltz writes several files;
-    if a3m is None:
+    # some are binary and crash its INPUT parser with KeyError '\x00'). Strip null bytes too.
+    candidates = sorted(out.rglob("*.a3m"), key=lambda p: p.stat().st_size, reverse=True)
+    chosen = None
+    for c in candidates:
+        raw = c.read_bytes()
+        if raw[:1] == b">" or b"\n>" in raw[:512]:
+            chosen = raw.replace(b"\x00", b"")
+            break
+    if chosen is None:
         weights.commit()
-        raise RuntimeError("MSA generation produced no .a3m")
+        raise RuntimeError(f"no valid a3m; candidates={[(str(p), p.stat().st_size) for p in candidates]}")
     msa_dir = Path(WEIGHTS) / "msa"
     msa_dir.mkdir(parents=True, exist_ok=True)
     dest = msa_dir / f"{label}.a3m"
-    shutil.copy(a3m, dest)
+    dest.write_bytes(chosen)
     weights.commit()
     return str(dest)