Disentangled Wasserstein Autoencoder for T-Cell Receptor Engineering

TL;DR

This paper introduces a disentangled Wasserstein autoencoder with an auxiliary classifier for protein sequence editing, specifically applied to T-cell receptors, enabling efficient and accurate functional modifications without altering structural backbone.

Contribution

It presents the first use of disentangled representations for TCR engineering, improving sequence editing efficiency and understanding of functional modifications.

Findings

Outperforms competing methods in generation quality

Requires only 10% of baseline models' running time

Enables one-pass protein sequence editing

Abstract

In protein biophysics, the separation between the functionally important residues (forming the active site or binding surface) and those that create the overall structure (the fold) is a well-established and fundamental concept. Identifying and modifying those functional sites is critical for protein engineering but computationally non-trivial, and requires significant domain knowledge. To automate this process from a data-driven perspective, we propose a disentangled Wasserstein autoencoder with an auxiliary classifier, which isolates the function-related patterns from the rest with theoretical guarantees. This enables one-pass protein sequence editing and improves the understanding of the resulting sequences and editing actions involved. To demonstrate its effectiveness, we apply it to T-cell receptors (TCRs), a well-studied structure-function case. We show that our method can be used to alter the function of TCRs without changing the structural backbone, outperforming several competing methods in generation quality and efficiency, and requiring only 10% of the running time needed by baseline models. To our knowledge, this is the first approach that utilizes disentangled representations for TCR engineering.