# Efficient Data Selection for Training Genomic Perturbation Models

arXiv: https://arxiv.org/abs/2503.14571 · HTML: https://mlsanity.com/paper/2503.14571/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 28 of its 56 references that are in the map.
- First posted on arXiv 18 Mar 2025.

## Builds on (most cited first, 28 of 28)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Simplifying Graph Convolutional Networks](https://mlsanity.com/paper/1902.07153/) (2019)
- [Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds](https://mlsanity.com/paper/1906.03671/) (2019)
- [BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning](https://mlsanity.com/paper/1906.08158/) (2019)
- [Active Learning on a Budget: Opposite Strategies Suit High and Low Budgets](https://mlsanity.com/paper/2202.02794/) (2022)
- [Unifying Graph Convolutional Neural Networks and Label Propagation](https://mlsanity.com/paper/2002.06755/) (2020)
- [Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training Data](https://mlsanity.com/paper/2108.01099/) (2021)
- [Subgroup Generalization and Fairness of Graph Neural Networks](https://mlsanity.com/paper/2106.15535/) (2021)
- [Bayesian Batch Active Learning as Sparse Subset Approximation](https://mlsanity.com/paper/1908.02144/) (2019)
- [Interventions, Where and How? Experimental Design for Causal Models at Scale](https://mlsanity.com/paper/2203.02016/) (2022)
- [A Framework and Benchmark for Deep Batch Active Learning for Regression](https://mlsanity.com/paper/2203.09410/) (2022)
- [Towards Understanding the Generalization of Graph Neural Networks](https://mlsanity.com/paper/2305.08048/) (2023)
- [Generalization in Graph Neural Networks: Improved PAC-Bayesian Bounds on Graph Diffusion](https://mlsanity.com/paper/2302.04451/) (2023)
- [Rethinking deep active learning: Using unlabeled data at model training](https://mlsanity.com/paper/1911.08177/) (2019)
- [Parting with Illusions about Deep Active Learning](https://mlsanity.com/paper/1912.05361/) (2019)
- [Modelling Cellular Perturbations with the Sparse Additive Mechanism Shift Variational Autoencoder](https://mlsanity.com/paper/2311.02794/) (2023)
- [Learning Causal Representations of Single Cells via Sparse Mechanism Shift Modeling](https://mlsanity.com/paper/2211.03553/) (2022)
- [Addressing the Item Cold-start Problem by Attribute-driven Active Learning](https://mlsanity.com/paper/1805.09023/) (2018)
- [Grain: Improving Data Efficiency of Graph Neural Networks via Diversified Influence Maximization](https://mlsanity.com/paper/2108.00219/) (2021)
- [RIM: Reliable Influence-based Active Learning on Graphs](https://mlsanity.com/paper/2110.14854/) (2021)
- [GeneDisco: A Benchmark for Experimental Design in Drug Discovery](https://mlsanity.com/paper/2110.11875/) (2021)
- [Near-Optimal Multi-Perturbation Experimental Design for Causal Structure Learning](https://mlsanity.com/paper/2105.14024/) (2021)
- [Information Gain Propagation: a new way to Graph Active Learning with Soft Labels](https://mlsanity.com/paper/2203.01093/) (2022)
- [DiscoBAX: Discovery of Optimal Intervention Sets in Genomic Experiment Design](https://mlsanity.com/paper/2312.04064/) (2023)
- [Uncertainty for Active Learning on Graphs](https://mlsanity.com/paper/2405.01462/) (2024)
- [Season combinatorial intervention predictions with Salt & Peper](https://mlsanity.com/paper/2404.16907/) (2024)
- [Neural Design for Genetic Perturbation Experiments](https://mlsanity.com/paper/2207.12805/) (2022)
- [Low-Regret Active learning](https://mlsanity.com/paper/2104.02822/) (2021)

## Abstract (arXiv)

Genomic studies face a vast hypothesis space, while interventions such as gene perturbations remain costly and time-consuming. To accelerate such experiments, gene perturbation models predict the transcriptional outcome of interventions. Since constructing the training set is challenging, active learning is often employed in a "lab-in-the-loop" process. While this strategy makes training more targeted, it is substantially slower, as it fails to exploit the inherent parallelizability of Perturb-seq experiments. Here, we focus on graph neural network-based gene perturbation models and propose a subset selection method that, unlike active learning, selects the training perturbations in one shot. Our method chooses the interventions that maximize the propagation of the supervision signal to the model. The selection criterion is defined over the input knowledge graph and is optimized with submodular maximization, ensuring a near-optimal guarantee. Experimental results across multiple datasets show that, in addition to providing months of acceleration compared to active learning, the method improves the stability of perturbation choices while maintaining competitive predictive accuracy.