Tranception: protein fitness prediction with autoregressive transformers and inference-time retrieval

by   Pascal Notin, et al.

The ability to accurately model the fitness landscape of protein sequences is critical to a wide range of applications, from quantifying the effects of human variants on disease likelihood, to predicting immune-escape mutations in viruses and designing novel biotherapeutic proteins. Deep generative models of protein sequences trained on multiple sequence alignments have been the most successful approaches so far to address these tasks. The performance of these methods is however contingent on the availability of sufficiently deep and diverse alignments for reliable training. Their potential scope is thus limited by the fact many protein families are hard, if not impossible, to align. Large language models trained on massive quantities of non-aligned protein sequences from diverse families address these problems and show potential to eventually bridge the performance gap. We introduce Tranception, a novel transformer architecture leveraging autoregressive predictions and retrieval of homologous sequences at inference to achieve state-of-the-art fitness prediction performance. Given its markedly higher performance on multiple mutants, robustness to shallow alignments and ability to score indels, our approach offers significant gain of scope over existing approaches. To enable more rigorous model testing across a broader range of protein families, we develop ProteinGym – an extensive set of multiplexed assays of variant effects, substantially increasing both the number and diversity of assays compared to existing benchmarks.


PoET: A generative model of protein families as sequences-of-sequences

Generative protein language models are a natural way to design new prote...

Few Shot Protein Generation

We present the MSA-to-protein transformer, a generative model of protein...

Guided Generative Protein Design using Regularized Transformers

The development of powerful natural language models have increased the a...

RITA: a Study on Scaling Up Generative Protein Sequence Models

In this work we introduce RITA: a suite of autoregressive generative mod...

Multi-level Protein Representation Learning for Blind Mutational Effect Prediction

Directed evolution plays an indispensable role in protein engineering th...

Interpretable Structured Learning with Sparse Gated Sequence Encoder for Protein-Protein Interaction Prediction

Predicting protein-protein interactions (PPIs) by learning informative r...

Sparse generative modeling of protein-sequence families

Pairwise Potts models (PM) provide accurate statistical models of famili...

Please sign up or login with your details

Forgot password? Click here to reset