Research map / Probabilistic and scientific modeling / Bayesian optimization
BioBO: Biology-informed Bayesian Optimization for Perturbation Design
Yanke Li, Tianyu Cui, Tommaso Mansi, Mangal Prakash, Rui Liao
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 |
|---|---|---|---|
| add Hallmark enrichment prior to UCB acquisition | UCB, Fusion embedding · GeneDisco IFN-gamma | cumulative top-k recall: 0.100 -> 0.109 | Improved |
| add Hallmark enrichment prior to UCB acquisition | UCB, Fusion embedding · GeneDisco IL-2 | cumulative top-k recall: 0.174 -> 0.178 | Improved |
| add GO enrichment prior to UCB acquisition | UCB, Fusion embedding · GeneDisco IL-2 | cumulative top-k recall: 0.174 -> 0.169 | Hurt |
| add enrichment prior to acquisition functions | UCB/EI/TS · GeneDisco (24 settings) | cumulative top-k recall: 23/24 cases | Improved |
| use multimodal Fusion embeddings instead of single modality (ablation) | BNN surrogate · GeneDisco IFN-gamma, IL-2 | labeling effort saving: 4% to 40% saving | Improved |
| use Fusion embeddings for surrogate (ablation) | BNN surrogate · GeneDisco | global log-likelihood | No clear effect |
| use Fusion embeddings for surrogate (ablation) | BNN surrogate · GeneDisco | log-likelihood near optimum | Improved |
| replace concatenation fusion with latent-space fusion (ablation) | BioBO · GeneDisco | BO performance | Improved |
| use DiscoBAX acquisition instead of standard AFs (ablation) | BO · GeneDisco | cumulative top-k recall | Hurt |
| use pure enrichment-analysis greedy selection instead of random (ablation) | EA only · GeneDisco IFN-gamma, IL-2 | cumulative top-k recall | Improved |
| use smaller beta (weaker enrichment prior) vs beta 1-5 (ablation) | BioBO · GeneDisco IFN-gamma | BO performance | Hurt |
| add enrichment prior to UCB with Achilles embedding | UCB · GeneDisco IFN-gamma | labeling efficiency: 20% | Improved |
Closest papers by meaning
- BioDiscoveryAgent: An AI Agent for Designing Genetic Perturbation Experiments (ICLR 2025 · similarity 0.86)
- Contextualizing biological perturbation experiments through language (ICLR 2025 · similarity 0.85)
- PRESCRIBE: Predicting Single-Cell Responses with Bayesian Estimation (NeurIPS 2025 · similarity 0.85)
- A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences (NeurIPS 2024 · similarity 0.85)
- Bayesian Optimization of Antibodies Informed by a Generative Model of Evolving Sequences (ICLR 2025 · similarity 0.85)
- A General Framework for User-Guided Bayesian Optimization (ICLR 2024 · 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.
