Research map / Probabilistic and scientific modeling / Bayesian optimization
Many Needles in a Haystack: Active Hit Discovery for Perturbation Experiments
Andrea Rubbi, Arpit Merchant, Samuel Ogden, Amir Akbarnejad, Pietro Lio, Sattar Vakili, Mohammad Lotfollahi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use Probability-of-Hit instead of Random | acquisition function · Schmidt IL-2 (b=200) | CHR@10 hits: 221.2 -> 575.4 | Improved |
| use Probability-of-Hit instead of Top-K Greedy | acquisition function · Schmidt IL-2 (b=200) | CHR@10 hits: 516.6 -> 575.4 | Improved |
| use Probability-of-Hit instead of Thompson Sampling | acquisition function · Sanchez Tau (b=200) | CHR@10 hits: 377.4 -> 389.4 | Improved |
| use Probability-of-Hit instead of Thompson-Hit | acquisition function · Zhuang NK (b=200) | CHR@10 hits: 252.4 -> 246.0 | Hurt |
| use Probability-of-Hit instead of Thompson Sampling | acquisition function · Zhu SARS-CoV-2 (b=200) | CHR@10 hits: 261.0 -> 251.6 | Hurt |
| use Probability-of-Hit instead of Top-K Greedy | acquisition function · Perturbation-SEM (6D) t=10 | cumulative hit ratio: 0.46 -> 0.49 | Improved |
| use Probability-of-Hit instead of Thompson-Hit | acquisition function · Perturbation-Pathways (4D) t=10 | cumulative hit ratio: 0.49 -> 0.62 | Improved |
| use Probability-of-Hit instead of Thompson Sampling | acquisition function · Perturbation-SEM (6D) t=10 | cumulative hit ratio: 0.3 -> 0.49 | Improved |
| use Probability-of-Hit instead of Top-K Greedy | acquisition function · Sine (2D) t=10 | cumulative hit ratio: 0.84 -> 0.88 | Improved |
| use Probability-of-Hit instead of Top-K Greedy | acquisition function · Schmidt IL-2 at 5 cycles (b=200) | hits recovered: 304.6 -> 336.6 | Improved |
| use Probability-of-Hit instead of Top-K Greedy | acquisition function · Zhu SARS-CoV-2 at 5 cycles (b=200) | hits recovered: 141.4 -> 137.6 | Hurt |
| increase batch size from b=2 to b=10 | Probability-of-Hit · Perturbation-Pathways (4D) | CHR@10: 4-fold improvement | Improved |
Closest papers by meaning
- Identifying biological perturbation targets through causal differential networks (ICML 2025 · similarity 0.86)
- PRESCRIBE: Predicting Single-Cell Responses with Bayesian Estimation (NeurIPS 2025 · similarity 0.85)
- BioDiscoveryAgent: An AI Agent for Designing Genetic Perturbation Experiments (ICLR 2025 · similarity 0.85)
- Contextualizing biological perturbation experiments through language (ICLR 2025 · similarity 0.85)
- PerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis (NeurIPS 2025 · similarity 0.85)
- Active Target Discovery under Uninformative Priors: The Power of Permanent and Transient Memory (NeurIPS 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
