Neural Design for Genetic Perturbation Experiments

TL;DR

This paper introduces a novel bandit optimization framework called OAE for efficiently exploring genetic perturbation spaces to maximize cellular phenotypes, addressing experimental limitations in drug development.

Contribution

It develops the OAE principle for batch query optimization, analyzes its convergence via Eluder dimension, and demonstrates superior performance in genetic and benchmark datasets.

Findings

OAE outperforms existing strategies in simulated and real datasets.

OAE achieves near-optimal solutions with fewer experiments.

OAE surpasses benchmarks in the GeneDisco challenge.

Abstract

The problem of how to genetically modify cells in order to maximize a certain cellular phenotype has taken center stage in drug development over the last few years (with, for example, genetically edited CAR-T, CAR-NK, and CAR-NKT cells entering cancer clinical trials). Exhausting the search space for all possible genetic edits (perturbations) or combinations thereof is infeasible due to cost and experimental limitations. This work provides a theoretically sound framework for iteratively exploring the space of perturbations in pooled batches in order to maximize a target phenotype under an experimental budget. Inspired by this application domain, we study the problem of batch query bandit optimization and introduce the Optimistic Arm Elimination ($\mathrm{OAE}$) principle designed to find an almost optimal arm under different functional relationships between the queries (arms) and the outputs (rewards). We analyze the convergence properties of $\mathrm{OAE}$ by relating it to the Eluder dimension of the algorithm's function class and validate that $\mathrm{OAE}$ outperforms other strategies in finding optimal actions in experiments on simulated problems, public datasets well-studied in bandit contexts, and in genetic perturbation datasets when the regression model is a deep neural network. OAE also outperforms the benchmark algorithms in 3 of 4 datasets in the GeneDisco experimental planning challenge.