SDD Wiki

PHGDH Autonomous Drug-Discovery Project — Plan v2

Last reorganized: 2026-04-18. Project target: novel allosteric/RBD-site PHGDH modulators for neurodegeneration (Alzheimer’s focus).

1. Architecture (3 tiers)

TierRoleSubstrate
0OS-level daemons (no LLM)LaunchDaemons + cron + scheduler_loop.sh
1Autonomous agentsjakeclaw (Gemma 4 26B @ L0), qwenclaw (Qwen 3.5 @ W0), Hermes
2Human consultantClaude — design + unstick only, minimize cost

Hosts: L0 = 192.168.68.200 (inference). W0 = 192.168.68.201 (agents, dashboard, scheduler). Cheaha = UAB HPC via SSH alias cheaha (user=jakechen).

2. 5 Aims

AimTitleStatusCompleted tasks
1Automated data pipeline + enrichment🟡 in progress1.1
2Computational analysis of PHGDH inhibitors✅ complete2.1, 2.2, 2.3
3Literature synthesis & knowledge graph🟡 in progress3.1
4Structure-based virtual screening on Cheaha🟡 active4.1
5Scientific deliverables (paper, slides, wiki)⚪ pending
Infrastructure (cross-cutting)✅ completeliveness, notifier, AGE, resilience, auto-resolver, plan-sync

3. PLEASER iteration loop (per task)

7 stages, single-char codes: P L E A S X R (X = eXpense; see STAGE_CODES.md). One active task at a time. AGE scored 1–9 reverse-NIH. Stage-timeout thresholds enforced by liveness → auto-resolver.

4. Infrastructure skills built (2026-04-18)

SkillPurpose
liveness-audit10-channel health monitor every 5 min
age-scoringAccuracy + Growth + Expense per iteration vs 7-day baseline
notifierTelegram alerts on milestones (dedup + rate-limit + queue)
resilienceAtomic writes, retry queues, hourly checkpoint, boot recovery, scheduler_loop
auto-resolverSelf-healing: per-channel playbook + 4-step escalation ladder
plan-sync(new) Auto-updates BACKLOG completion status from COMPLETED + artifacts

5. Data assets (on W0 ~/.openclaw/workspace/data/)

  • aim2_top20_inhibitors.csv — 20 rows, pv ≥ 7, ranked
  • aim2_pdb_binding_sites.json — 5 Å residue shells per PDB ligand
  • aim3_structures/ — 4NJN, 6RIH, 6PLF, 7S3R, 2G76 (PDB files)
  • aim3_pubmed_phgdh_allosteric.json — 29 papers 2024–2026
  • aim4_smiles/smiles_potency_sorted.tsv — 1011 molecules
  • aim4_docking_results/ — Vina dock poses (currently 1; scaling to 20)

6. Key decisions & conventions

  • Publish gate: PUBLISH-01 — Jake triggers publish, daemons never push
  • Stage codes: never rename X (eXpense)
  • Scheduling: macOS SSH can’t install LaunchAgents/cron → scheduler_loop nohup with 3 resurrection paths
  • IPs: L0=.200, W0=.201 (NOT .133/.134)
  • Cheaha: SSH alias cheaha as user jakechen, key id_ed25519, no VPN for login node

7. Next 48 h

  1. Task 4.2 — scale Vina to top-20 inhibitors × 6RIH K4T pocket
  2. Task 1.2 — Hermes data→knowledge enrichment loop
  3. Task 5.2 — Repo README + pipeline diagram
  4. Move L2 report cron 17:00 → 23:00 (needs user-terminal sudo)
  5. User-gated publish of 3 wiki drafts + 20 concept-hub stubs

8. Automatic plan maintenance

Starting 2026-04-18, plan-sync runs:

  • Every hour via scheduler_loop — scans COMPLETED.md + artifact files, auto-updates BACKLOG.md task statuses (pending → DONE when evidence found)
  • At every R-stage exit — post_r_watchdog calls plan-sync before promoting
  • Writes project-state/PLAN_STATUS.md — a one-page human-readable dashboard
  • Telegram 📋 notice when any status flip occurs

No more stale plans.