A Pareto-optimal compositional energy-based model for sampling and optimization of protein sequences
This work addresses the problem of designing protein sequences with multiple desired properties for researchers in computational biology and drug discovery, representing a novel method for a known bottleneck rather than an incremental improvement.
The paper tackled the challenge of multi-objective optimization in protein sequence design, where properties are independent or orthogonal, by proposing a Pareto-compositional energy-based model (pcEBM) that uses multiple gradient descent to sample sequences adhering to various constraints, and demonstrated its ability to learn non-convex Pareto fronts and generate sequences satisfying multiple desired properties in real-world antibody design tasks.
Deep generative models have emerged as a popular machine learning-based approach for inverse design problems in the life sciences. However, these problems often require sampling new designs that satisfy multiple properties of interest in addition to learning the data distribution. This multi-objective optimization becomes more challenging when properties are independent or orthogonal to each other. In this work, we propose a Pareto-compositional energy-based model (pcEBM), a framework that uses multiple gradient descent for sampling new designs that adhere to various constraints in optimizing distinct properties. We demonstrate its ability to learn non-convex Pareto fronts and generate sequences that simultaneously satisfy multiple desired properties across a series of real-world antibody design tasks.