Jun 3 07:01 docs: update citation with live bioRxiv DOI 10.64898/2026.05.29.728379 May 20 13:18 refresh: rename class_names inactive -> background in eval JSONs May 6 11:39 Remove stale connection_type_coverage.json (no longer generated by candidate analysis pipeline) 11:39 Refresh figure data: drop Archived/, refresh Candidate_Analysis with rescored hits Removed 56 files in Generation_Model/End2end/Archived/ (legacy single-run experiments, not used by current figures). Removed obsolete Candidate_Analysis/AMP_{active_druglike,potent_druglike}.csv, replaced by AMP_active.csv (39,891 active candidates) + AMP_hit.csv (799 hits, triple filter) generated after the AMP ensemble retrain on MIC-only DBAASP. All other per-stage and end-to-end metrics re-uploaded; LFS dedup keeps unchanged content. Apr 7 16:01 docs: UTF-8 author name 15:58 docs: bioRxiv citation with DOI placeholder 15:52 fix: use @misc 15:49 docs: update title and citation to bioRxiv preprint format 15:01 docs: update Corresponding Figures table and fix Connection dir name Apr 5 14:59 data: update Candidate_Analysis (AMP_potent→AMP_active rename, updated eval_report) 14:58 docs: update README — rename Potent→Active, fix ratios, remove connection_type_coverage Apr 1 18:54 docs: add dataset card with file structure, figure mapping, and quick start 18:51 chore: upload raw.jsonl replacing raw.json (Flat/run5) 18:51 chore: upload raw.jsonl replacing raw.json (Flat/run4) 18:51 chore: upload raw.jsonl replacing raw.json (Flat/run3) 18:51 chore: upload raw.jsonl replacing raw.json (Flat/run2) 18:51 chore: upload raw.jsonl replacing raw.json (Flat/run1) 18:51 chore: remove old raw.json (Flat/run5) 18:51 chore: remove old raw.json (Flat/run4) 18:51 chore: remove old raw.json (Flat/run3) 18:51 chore: remove old raw.json (Flat/run2) 18:51 chore: remove old raw.json (Flat/run1) 18:51 chore: update eval_report.json with per_type_nn_distance (Flat/run5) 18:51 chore: update eval_report.json with per_type_nn_distance (Flat/run4) 18:51 chore: update eval_report.json with per_type_nn_distance (Flat/run3) 18:51 chore: update eval_report.json with per_type_nn_distance (Flat/run2) 18:51 chore: update eval_report.json with per_type_nn_distance (Flat/run1) 18:51 chore: update eval_report.json with per_type_nn_distance (Cascade/run5) 18:50 chore: update eval_report.json with per_type_nn_distance (Cascade/run4) 18:50 chore: update eval_report.json with per_type_nn_distance (Cascade/run3) 18:50 chore: update eval_report.json with per_type_nn_distance (Cascade/run2) 18:50 chore: update eval_report.json with per_type_nn_distance (Cascade/run1) 13:37 chore: remove script from Data directory (run_comparison.py) 13:37 chore: remove script from Data directory (prepare_constraints.py) 13:37 chore: remove script from Data directory (Infilling_Generation.ipynb) 13:37 chore: remove script from Data directory (constrained_generation.ipynb) 13:37 chore: remove script from Data directory (flat_cascade_multirun.ipynb) Mar 16 14:36 refactor: convert raw.json to raw.jsonl for consistency 10:38 Add pre-computed connection type coverage stats 10:05 Remove Archived directory (not needed for reproduction) 10:04 Restructure: all figure data under Data/ prefix 10:02 Delete batch 4 (73 files) 10:02 Delete batch 3 (200 files) 10:02 Delete batch 2 (200 files) 10:02 Delete batch 1 (200 files) 10:00 Restructure: all figure data under Data/ prefix Mar 13 12:10 Add eval data (eval_test.json + metrics.json) for paper figure reproduction Mar 11 16:16 Upload README.md with huggingface_hub 16:10 Upload Candidate_Analysis/eval_report.json with huggingface_hub 16:10 Upload Candidate_Analysis/AMP_potent_druglike.csv with huggingface_hub 
