Abstract:Deep generative models (DGMs) have shown great success in the understanding and data-driven design of proteins. Variational autoencoders (VAEs) are a popular DGM approach that can learn the correlated patterns of amino acid mutations within a multiple sequence alignment (MSA) of protein sequences and distill this information into a low-dimensional latent space to expose phylogenetic and functional relationships and guide generative protein design. Autoregressive (AR) models are another popular DGM approach that typically lacks a low-dimensional latent embedding but does not require training sequences to be aligned into an MSA and enable the design of variable length proteins. In this work, we propose ProtWave-VAE as a novel and lightweight DGM, employing an information maximizing VAE with a dilated convolution encoder and an autoregressive WaveNet decoder. This architecture blends the strengths of the VAE and AR paradigms in enabling training over unaligned sequence data and the conditional generative design of variable length sequences from an interpretable, low-dimensional learned latent space. We evaluated the model's ability to infer patterns and design rules within alignment-free homologous protein family sequences and to design novel synthetic proteins in four diverse protein families. We show that our model can infer meaningful functional and phylogenetic embeddings within latent spaces and make highly accurate predictions within semisupervised downstream fitness prediction tasks. In an application to the C-terminal SH3 domain in the Sho1 transmembrane osmosensing receptor in baker's yeast, we subject ProtWave-VAE-designed sequences to experimental gene synthesis and select-seq assays for the osmosensing function to show that the model enables synthetic protein design, conditional C-terminus diversification, and engineering of the osmosensing function into SH3 paralogues.

Robust Model-Based Optimization for Challenging Fitness Landscapes

Global-Context Aware Generative Protein Design

ProteinInvBench: Benchmarking Protein Inverse Folding on Diverse Tasks, Models, and Metrics.

Generative De Novo Protein Design with Global Context

Robust Optimization in Protein Fitness Landscapes Using Reinforcement Learning in Latent Space

Improving Protein Optimization with Smoothed Fitness Landscapes

Computational and Experimental Exploration of Protein Fitness Landscapes: Navigating Smooth and Rugged Terrains

FLOP: Tasks for Fitness Landscapes Of Protein wildtypes

Protein Design by Integrating Machine Learning with Quantum Annealing and Quantum-inspired Optimization

Multi-Scale Representation Learning for Protein Fitness Prediction

Protein Sequence Design with Batch Bayesian Optimisation

Designing diverse and high-performance proteins with a large language model in the loop

Proximal Exploration for Model-guided Protein Sequence Design

Protein Language Model Supervised Precise and Efficient Protein Backbone Design Method

Optimistic Games for Combinatorial Bayesian Optimization with Application to Protein Design

Learning protein fitness models from evolutionary and assay-labeled data

Robust Optimization Using a Methodology Based on Cross Entropy Methods

An integrative approach to protein sequence design through multiobjective optimization

Multi-Objective Latent Space Optimization of Generative Molecular Design Models

Protein Design by Directed Evolution Guided by Large Language Models

ProtWave-VAE: Integrating Autoregressive Sampling with Latent-Based Inference for Data-Driven Protein Design