0.08 correlation on million-dollar datasets: researchers question virtual cell capital allocation
anshulkundaje · x · 2026-10-06
A widely shared critique points out that predicted perturbation effects in virtual cell efforts show only a 0.08 Pearson correlation with observations — on datasets that cost millions to generate — which the author argues should force a rethink of capital allocation.
Resharing it, suragnair lists what the field actually needs for useful virtual cells: non-destructive sampling of the same cell at <30-min intervals with high-quality measurements, protein measurements and localization, TF-DNA binding data, and more. The takeaway: expensive data generation with weak model validation means measurement infrastructure and benchmarks must come before more models.
More from Research
- Cornell Tech prof presents two pretraining-behavior papers at COLM, recruiting PhDs and postdocs — pratyushmaini · 2026-10-06
- Non-Expert Runs AI-Driven Interpretability Study Claiming Linear Representation of Directive Force in LLMs — LoudYogurtcloset7856 · 2026-10-06
- AC2 lets LLM RL train faster than GRPO by trusting critics on partial rollouts — stanfordnlp · 2026-10-06
- Stanford CS224N Winter 2026 posts free slides and assignments online — stanfordnlp · 2026-10-06
- Only right answers still teach chemistry: popular chemistry benchmark has shortcuts — tak3sh8 · 2026-10-06
- UT Austin math chair: OpenAI appears set to release ~400 AI-generated proofs at once — 141_1337 · 2026-10-06