On generative models of T-cell receptor sequences

TL;DR

This paper compares knowledge-guided and deep learning models for predicting T-cell receptor sequence distributions, finding the former more accurate, interpretable, and computationally efficient.

Contribution

It introduces and evaluates a knowledge-guided, physics-inspired model against a deep neural network for TCR sequence modeling, highlighting advantages of the former.

Findings

Knowledge-guided model outperforms deep network in prediction accuracy.

Knowledge-guided model is more interpretable.

Knowledge-guided model requires less computational resources.

Abstract

T-cell receptors (TCR) are key proteins of the adaptive immune system, generated randomly in each individual, whose diversity underlies our ability to recognize infections and malignancies. Modeling the distribution of TCR sequences is of key importance for immunology and medical applications. Here, we compare two inference methods trained on high-throughput sequencing data: a knowledge-guided approach, which accounts for the details of sequence generation, supplemented by a physics-inspired model of selection; and a knowledge-free Variational Auto-Encoder based on deep artificial neural networks. We show that the knowledge-guided model outperforms the deep network approach at predicting TCR probabilities, while being more interpretable, at a lower computational cost.