[cschanhniem]

Summary

• Updates CHANGELOG.md to document PRs docs: Stage 3 serum stability update with actual pilot-panel data #68 and docs: update synthesis readiness report for current 20-candidate pilot panel #69
• Locally regenerates synthesis readiness report, QC outputs, and diversity analysis for the current 20-candidate pilot panel (files are gitignored; updated for local wet-lab reference)
• Stale references to old 16-candidate confident_panel.csv (pre-PR fix: resolve CRITICAL pre-wetlab pipeline issues (SEED-008 excluded, benchmark config mismatch) #61 pipeline) corrected throughout

Key changes in synthesis_readiness_report.md (local file)

Field	Old	New
Panel	16-candidate confident_panel.csv (pre-PR #61)	20-candidate pilot_panel.csv (post PRs #61–68)
AUROC	0.8037	0.8047
Benchmark	44 AMPs	43 AMPs (post-dedup PR #66)
CI95%	[0.7087, 0.8895]	[0.7072, 0.8922]
External predictor gate	✅ PASS (16/16 old candidates)	⏳ PENDING (new candidates not submitted)
Safety range	0.525–1.000	0.845–1.000 (all candidates safer)
Diversity score	0.601 (4 clusters)	0.712 (6 clusters, 1 per seed family)
Minimal panel	4-candidate	6-candidate (1 per mechanism class)

New QC findings for pilot panel

• SEED-007 (5 candidates): HIGH SPPS difficulty due to Met + 3 trypsin sites — synthesisable but discuss with vendor
• SEED-008_VAR_051 (FPVTYRWWKWWKG): 4 Trp → photolability risk, amber vials required
• SEED-007_VAR_001/018/035/022: No UV chromophore (no Trp/Phe/Tyr) → use BCA/Bradford for concentration
• N-terminal acetylation (Ac-) recommended for ALL 20 candidates

Test plan

• 1200 tests passing (CHANGELOG only; output files are gitignored)
• Reviewer: confirm CHANGELOG entries for PR docs: Stage 3 serum stability update with actual pilot-panel data #68 and docs: update synthesis readiness report for current 20-candidate pilot panel #69 are accurate
• Wet-lab team: submit outputs/pilot_panel.fasta to external predictors before synthesis

🤖 Generated with Claude Code