Deploy c0f4b8a

Hygiene sweep: dead imports, dead locals, cosmetic f-strings, one stale type-hint

Source: https://github.com/WINTER4000/turingDNA/commit/c0f4b8a74488cfda0cd9a324c2a6ddc8c2ca2660

dee/auth.py

@@ -43,7 +43,6 @@ from __future__ import annotations
 import hashlib
 import hmac
 import json
-import urllib.parse
 import logging
 import os
 import threading
@@ -51,7 +50,7 @@ import time
 from dataclasses import dataclass
 from typing import Any, Dict, Optional
 
-from flask import Flask, Response, g, jsonify, make_response, request
+from flask import Flask, Response, g, jsonify, request
 
 logger = logging.getLogger("dee.auth")
 

dee/core/agent_tools.py

@@ -30,7 +30,7 @@ from __future__ import annotations
 import contextvars
 import logging
 import re
-from typing import Any, Callable, Dict, List, Optional
+from typing import Any, Dict, List, Optional
 
 logger = logging.getLogger("dee.agent_tools")
 
@@ -723,7 +723,7 @@ def _tool_design_primers(args: Dict[str, Any]) -> Dict[str, Any]:
 
     try:
         result = _P.design_primers(template, _oi(args.get("target_start")), _oi(args.get("target_end")))
-    except Exception as exc:  # noqa: BLE001 — mirrors the REST route's own catch-all
+    except Exception:  # noqa: BLE001 — mirrors the REST route's own catch-all
         logger.exception("design_primers tool call failed")
         return {"ok": False, "error": "Primer design failed — check the inputs."}
     if not result.get("ok"):
@@ -969,7 +969,7 @@ def _tool_design_variant_library(args: Dict[str, Any]) -> Dict[str, Any]:
         return {"ok": False, "kind": "busy", "error": (
             "The design engine is busy scoring another request right now — try again in a few seconds."
         )}
-    except (ImportError, OSError) as exc:
+    except (ImportError, OSError):
         # The scorer itself couldn't start — torch missing, model weights not
         # downloaded, no disk, OOM loading. Nothing to do with the sequence,
         # so don't tell the scientist to go check it. Same disease as the

dee/core/benchmark.py

@@ -21,7 +21,7 @@ from __future__ import annotations
 
 import re
 from dataclasses import dataclass
-from typing import Dict, List, Optional, Sequence
+from typing import List, Optional, Sequence
 
 import numpy as np
 

dee/core/cloning.py

@@ -181,7 +181,7 @@ def gibson(fragments: List[Dict[str, Any]], *, circular: bool = True,
 
     primers = _gibson_primers(order, circular, design_overlap, min_overlap, max_overlap)
     # carry fragment spans as features on the assembled product
-    feats, pos = [], 0
+    feats = []
     palette = ["#6A89E4", "#5FA98A", "#E4B45F", "#B07ED9", "#C77E7E", "#D98C5F"]
     cursor = 0
     for i in range(n):

dee/core/crispr.py

@@ -405,7 +405,7 @@ class Guide:
                                         # highlight on the AlphaFold structure.
 
 
-def guide_to_dict(g: "GuideRNA") -> Dict[str, Any]:
+def guide_to_dict(g: "Guide") -> Dict[str, Any]:
     """Every field the CRISPR results table reads, in one place.
 
     POST /api/crispr/design and the agent's design_crispr_guides tool both
@@ -924,15 +924,14 @@ def _predict_indels(
     # roughly to 100.
     top_n  = distribution[:_INDEL_TOP_N]
     rest_w = sum(w for (_, w, _) in distribution[_INDEL_TOP_N:])
-    rest_fs_w = sum(w for (_, w, fs) in distribution[_INDEL_TOP_N:] if fs)
     predicted: List[Tuple[str, float]] = [(lbl, w) for (lbl, w, _) in top_n]
     if rest_w > 0.001:
         predicted.append(("other", rest_w))
 
     # Frameshift % across the FULL distribution (not just top N), since
-    # KO probability depends on total non-multiple-of-3 fraction.
+    # KO probability depends on total non-multiple-of-3 fraction — the
+    # tail beyond top_n is included here, unlike `predicted` above.
     fs_total = sum(w for (_, w, fs) in distribution if fs) + 0.0
-    # rest_fs_w already included in fs_total via the `distribution` loop
 
     # Top-outcome dominance: the SINGLE most-likely repair outcome's
     # share of all events. Real biological signal: a guide where the

dee/core/crispr_cloning.py

@@ -40,7 +40,7 @@ References:
 from __future__ import annotations
 
 from dataclasses import dataclass
-from typing import Dict, List, Optional
+from typing import Dict, List
 
 
 @dataclass

dee/core/domains.py

@@ -222,8 +222,8 @@ def domains(accession: str, positions: Optional[Iterable] = None) -> Dict[str, A
         "source": "InterPro (EMBL-EBI)",
         "summary": _summarize(acc, entries, architecture, protein_length),
         "numbering": (
-            f"Positions are UniProt's, counted from residue 1 of the full "
-            f"precursor"
+            "Positions are UniProt's, counted from residue 1 of the full "
+            "precursor"
             + (f" ({protein_length} aa)" if protein_length else "")
             + ", including any signal peptide or propeptide. If the user is "
               "working from mature-protein numbering, every boundary here is "

dee/core/exon.py

@@ -36,7 +36,7 @@ import threading
 import time
 import urllib.error
 import urllib.request
-from dataclasses import dataclass, field
+from dataclasses import dataclass
 from typing import Dict, List, Optional, Tuple
 
 logger = logging.getLogger("dee.exon")

dee/core/harvest.py

@@ -62,7 +62,7 @@ import re
 import time
 import urllib.error
 import urllib.request
-from typing import Any, Dict, Iterable, List, Optional, Sequence, Tuple
+from typing import Any, Dict, List, Optional, Sequence, Tuple
 
 from Bio.Data.IUPACData import protein_letters_3to1
 

dee/core/mutagenesis.py

@@ -35,7 +35,7 @@ Which one a lab uses is a kit decision, so the caller says.
 from __future__ import annotations
 
 import re
-from typing import Any, Dict, List, Optional
+from typing import Any, Dict, Optional
 
 from dee.core.primers import gc_clamp, gc_percent, hairpin_score, revcomp
 from dee.core.primers import self_dimer_3p, tm_c

dee/core/orchestrator.py

@@ -2560,10 +2560,10 @@ def _model_note(model: str) -> str:
     lines = [
         "SCORING MODEL. Substitution scores on this workbench come from ESM-2, "
         "and the two sizes carry product names:",
-        f"  • ACHILLES 1.0 — ESM-2 35M. Runs on the workbench's own CPU. "
-        f"Seconds per protein, free, always available.",
-        f"  • PROMETHEUS 1.0 — ESM-2 650M. Runs on a rented GPU, so the first "
-        f"call waits for a container to start. Larger, and the same method.",
+        "  • ACHILLES 1.0 — ESM-2 35M. Runs on the workbench's own CPU. "
+        "Seconds per protein, free, always available.",
+        "  • PROMETHEUS 1.0 — ESM-2 650M. Runs on a rented GPU, so the first "
+        "call waits for a container to start. Larger, and the same method.",
         f"This run is using {active}. Say so plainly if asked, and name the "
         "underlying checkpoint too — the product name is a label on a public "
         "model, not a proprietary one, and implying otherwise would overstate "

dee/core/outcomes.py

@@ -25,7 +25,7 @@ numpy only; runs in milliseconds on CPU.
 from __future__ import annotations
 
 import datetime as _dt
-from dataclasses import dataclass, field
+from dataclasses import dataclass
 from typing import Callable, Dict, List, Optional, Sequence, Tuple
 
 import numpy as np

dee/core/remote_scorer.py

@@ -19,7 +19,7 @@ instead of quietly working badly.
 from __future__ import annotations
 
 import logging
-from typing import Any, Dict, List, Optional
+from typing import Dict, List
 
 import pandas as pd
 

dee/core/resolution_cache.py

@@ -45,7 +45,7 @@ import re
 import threading
 import time
 from collections import OrderedDict
-from typing import Any, Dict, Optional, Tuple
+from typing import Any, Dict, Optional
 
 # Entries older than this are re-fetched. Sequence records are stable over
 # weeks; this is about eventually noticing a re-annotation, not about

dee/core/voice_vocab.py

@@ -37,7 +37,7 @@ Nothing here commits the product to a vendor.
 from __future__ import annotations
 
 import re
-from typing import Any, Dict, Iterable, List, Optional
+from typing import Iterable, List, Optional
 
 # Vendor guidance converges on a few hundred; past that, accuracy on ordinary
 # speech starts to suffer as boosted terms pull neighbours toward them.

dee/optimizer/search.py

@@ -25,7 +25,6 @@ import random
 from dataclasses import dataclass, field
 from typing import Dict, FrozenSet, List, Optional, Tuple
 
-import numpy as np
 import pandas as pd
 
 logger = logging.getLogger(__name__)

dee/server.py

@@ -28,7 +28,6 @@ import re
 import tempfile
 import threading
 import time
-import traceback
 import uuid
 from collections import deque
 from dataclasses import dataclass, field
@@ -54,8 +53,8 @@ from dee.core.sequence import (
     list_cds_features,
     parse_input,
 )
-from dee.models.scorer import ESM2Scorer, ScorerConfig, top_percentile_pool
-from dee.optimizer.search import SearchConfig, apply_variant, evolve
+from dee.models.scorer import top_percentile_pool
+from dee.optimizer.search import SearchConfig, evolve
 
 logger = logging.getLogger("dee.server")
 
@@ -3809,7 +3808,7 @@ def create_app() -> Flask:
             "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
         )
         resp.headers["Content-Disposition"] = (
-            f'attachment; filename="turingdna_crispr_guides.xlsx"'
+            'attachment; filename="turingdna_crispr_guides.xlsx"'
         )
         return resp
 
@@ -4258,7 +4257,6 @@ def _compute_de_ll_maps(grouped_with_lib):
     library) → only the opt-in rebuild calls it. Any failure for a library
     yields None for that unit (its keys stay unbinned). Returns a list parallel
     to grouped_with_lib."""
-    from dee.models.scorer import ESM2Scorer, ScorerConfig
     cache: Dict[str, Any] = {}
     scorer = None
     maps = []