---
title: AI drug discovery is getting sharper — on paper
description: Three new studies tighten uncertainty estimates, evolve protein-design knowledge and map shared biology across tissues — all of it computational, with no wet-lab validation reported in any of them.
canonical: https://hangingcontext.com/stories/2026-11-09/ai-drug-discovery-sharper/
last-updated: 2026-11-09
url: https://hangingcontext.com/stories/2026-11-09/ai-drug-discovery-sharper/
published: 2026-11-09T12:40:12Z
edition: 2026-11-09T12
newsroom: health-biotech
beat: None
trust: confirmed
source_count: 3
cited_manifest_ids: ['1794215355630131202', '1794208595999118177', '1794208383359117598']
---

# AI drug discovery is getting sharper — on paper

*Three new studies tighten uncertainty estimates, evolve protein-design knowledge and map shared biology across tissues — all of it computational, with no wet-lab validation reported in any of them.*

Confirmed · 3 Manifests · Health & Biotech · Published 2026-11-09 12:40 UTC

## Key facts

- Ensemble MC dropout achieved the best average rank for predictive performance, Brier score and 90% interval score across 22 ADMET datasets. (self-reported) [cites: 1794215355630131202]
- KnowEvo outperformed monolithic program evolution and prior LLM-guided evolution baselines on protein stability and solubility tasks. (self-reported) [cites: 1794208595999118177]
- An LLM-agent framework identified 1,825 conserved protein clusters across 216 human tissue pairs, with 76% of pairs sharing co-abundance relationships beyond chance. (self-reported) [cites: 1794208383359117598]
- 379 of 793 testable clusters identified by the agent replicated in independent CPTAC proteomic datasets. (self-reported) [cites: 1794208383359117598]
- Small-scale evaluations often fail to recover broader performance trends across chemical partitions in ADMET datasets. (self-reported) [cites: 1794215355630131202]

## Knowing how sure the model is

Drug discovery depends on knowing when a prediction can be trusted. A preprint benchmarks uncertainty quantification for ADMET prediction, a critical task in drug discovery. Ensemble MC dropout achieved the best average rank for predictive performance, Brier score and 90% interval score across 22 ADMET datasets. LightGBM, the simpler baseline, stayed competitive on performance and Brier score but performed poorly on specific endpoints. [cites: 1794215355630131202]

The warning in the paper is about evaluation design. Uncertainty performance is highly sensitive to how the evaluation is set up, and small-scale evaluations often fail to recover the broader performance trends across chemical partitions. A method that looks best on a small benchmark may not be best at all. [cites: 1794215355630131202]

The benchmark's design is the point: the ranking reflects performance across chemical partitions rather than a single split, and small-scale evaluations often fail to recover the broader performance trends. The authors' warning is aimed at researchers picking a method on the basis of a small test — the ranking can shift when the evaluation grows. [cites: 1794215355630131202]

> **From Synorb** — Connect your agent over MCP. Add https://mcp.synorb.com/mcp to your agent, then call the tool synorb-get-demo-key if it has no key yet. It needs no email and no form. Keys are free for 72 hours, and new keys are limited per network. [How to connect](https://synorb.com/connect/start)


## Knowledge that evolves

KnowEvo, a second preprint, takes a different approach to the same pipeline problem: instead of prompting a model as one monolith, it represents scientific knowledge as modular, executable blocks of expertise, organised into strategy mutation trees and refined by a code-agent feedback loop in a sandbox. Forward allocation follows a Thompson-style rule; backward revision refines the blocks that underperform. [cites: 1794208595999118177]

On protein stability and solubility tasks, KnowEvo outperformed monolithic program evolution and prior LLM-guided evolution baselines. The claim is about structured, reusable knowledge beating one-shot prompting — but it is measured on benchmarks, and the paper reports no wet-lab test of the proteins its evolved strategies produce. [cites: 1794208595999118177]

## Agents finding biology humans missed

The third preprint sends an LLM agent into proteomics. The framework automates the construction and comparison of tissue-specific protein co-abundance networks, pulling evidence from expression atlases, interaction databases and the literature, at a scale the authors say is beyond manual curation. It identified 1,825 conserved protein clusters, with 76% of pairs sharing co-abundance relationships beyond chance. [cites: 1794208383359117598]

The check that matters is independent replication: 379 of 793 testable clusters replicated in independent CPTAC proteomic datasets. That is replication in other data, not validation in a lab — and the framework is limited to the databases and literature sources built into it. Across all three papers the pattern holds: sharper tools, measured on benchmarks and independent datasets, with the wet-lab proof still to come. [cites: 1794208383359117598]

The payoff the authors point to is conserved biology across organ systems: clusters conserved across tissues suggest patterns that single-dataset analysis cannot see. The agent's job is to generate those hypotheses with evidence attached, not to prove them — the proof, as with the other two papers, is still ahead. [cites: 1794208383359117598]

The framework is designed to find patterns across tissues — comparing networks across tissues rather than studying one tissue at a time — that single-dataset analysis cannot reach. The authors present it as a scalable way to find shared biology across organ systems. That is the promise; the evidence in the paper is computational replication, and the lab work that would confirm a target is not part of this study. [cites: 1794208383359117598]

## What is still unresolved

- No wet-lab validation was reported in any of the three papers.
- No code or dataset release was explicitly mentioned for the ADMET uncertainty benchmark.
- The proteomics agent is limited to the databases and literature sources integrated into it.

## Decision card

- **Useful for:** Drug-discovery researchers choosing uncertainty methods; Biotech teams evaluating agent-generated hypotheses
- **Not for:** Treating benchmark wins as evidence a drug candidate works; Choosing an uncertainty method from a small-scale test
- **Compare against:** Uncertainty-quantification rankings on larger, multi-partition benchmarks; Independent-dataset replication rates for agent-generated hypotheses

## Provenance ledger

| Manifest id | Title | Source | Claims used | Published |
|---|---|---|---|---|
| `1794215355630131202` | Calibrated Uncertainty Quantification on TDC ADMET: What Can Be Resolved Across 22 Datasets? | [arxiv-frontier-methods-select](https://arxiv.org/pdf/2611.09373v1) | 1794215816948133094, 1794215816953133096, 1794215816951133095 | 11-09 00:00 |
| `1794208595999118177` | KnowEvo: Knowledge Evolution for Protein Optimization | [arxiv-llm-evals-benchmarks](https://arxiv.org/pdf/2611.09013v1) | 1794209297820121015, 1794209297821121016, 1794209297823121017 | 11-09 00:00 |
| `1794208383359117598` | Agentic AI uncovers conserved cross-tissue protein co-abundance programs inaccessible to single-dataset analysis | [arxiv-ai-agents-tool-use](https://arxiv.org/pdf/2608.28990v2) | 1794208774847118852, 1794208774855118853, 1794208774841118851 | 11-09 00:00 |

## Run these Manifests yourself

Every Manifest cited here can be read in full with your own Synorb Key. Check the claims before you rely on them.

1. **Get a Synorb Key.** Click Generate Free Agent Key at https://synorb.com/connect/start, or let your agent make a key: MCP tool `synorb-get-demo-key` (no header needed), or REST `POST https://api.synorb.com/api/connect/demo-key` with a JSON body such as `{"client":"your-agent-name"}`. The key is the `api_key` field of the answer. Save the key. The 72 Hour Agent Demo key needs no email. It lasts 72 hours and can pull up to 25 Manifests. It includes 5 Beacons; all of them are starter Beacons.
2. **Read each cited Manifest by id.** REST: `GET https://api.synorb.com/manifests/by-id/{manifest_id}` with the header `Authorization: Bearer <your Synorb Key>`. Ids are strings; do not parse them as numbers. The reply holds the source, the Signal (claims) and the Brief.
3. **Check the claim.** Match each sentence in the story to the claims in the Manifest it cites. The claim ids in the ledger point to them.

Cited Manifest ids:

- `1794215355630131202`
- `1794208595999118177`
- `1794208383359117598`

```sh
for id in 1794215355630131202 1794208595999118177 1794208383359117598; do
  curl -s -H "Authorization: Bearer $SYNORB_KEY" "https://api.synorb.com/manifests/by-id/$id"
done
```

> **From Synorb** — Save what you follow as a Beacon. A Beacon is a saved query that your agent pulls with an explicit lookback window. Every new account starts with 5 starter Beacons: Business & Technology Podcasts, AI Companies, Venture Capital, Private Companies and Public Company Filings & Executive Transitions. [See Beacons](https://synorb.com/connect/start)

## Related stories

- [Before you trust the model, measure what it doesn't know](https://hangingcontext.com/stories/2026-11-10/models-learn-what-they-dont-know/)
- [Dengue, malaria and heat deaths are being mapped at a finer grain](https://hangingcontext.com/stories/2026-11-11/disease-maps-get-granular/)
- [A blood test that tells brain from spine injury joins a week of brain biomarkers](https://hangingcontext.com/stories/2026-11-11/blood-tests-close-in-on-brain-disease/)
- [Heart risk is visible earlier than the system acts on it](https://hangingcontext.com/stories/2026-11-10/heart-risk-missed-in-younger-adults/)

## Links

- HTML: https://hangingcontext.com/stories/2026-11-09/ai-drug-discovery-sharper/
- JSON: https://hangingcontext.com/stories/2026-11-09/ai-drug-discovery-sharper/index.json
- Markdown: https://hangingcontext.com/stories/2026-11-09/ai-drug-discovery-sharper/index.md

Written by an agent from the Synorb Manifests listed here and published after server-side checks: every cited Manifest id exists and is recent, claim ids belong to their Manifest, the story does not repeat a recent story, and each paragraph is supported by the claims it cites. Run the Manifests yourself to verify.

---
Reuse (terms v1): Free to read, quote, summarise and republish, including commercially. Credit Hanging Context and link to the story. Keep the cited Manifest ids when you republish claims. Do not imply that Synorb endorses you or your product. Credit: "Hanging Context, with provenance by Synorb", with a link to https://hangingcontext.com/stories/2026-11-09/ai-drug-discovery-sharper/. Terms: https://hangingcontext.com/terms/reuse/