Improving Protein Optimization with Smoothed Fitness Landscapes
Preprint
Abstract
The ability to engineer novel proteins with higher fitness for a desired property would be revolutionary for biotechnology and medicine. Modeling the combinatorially large space of sequences is infeasible; prior methods often constrain optimization to a small mutational radius, but this drastically limits the design space. Instead of heuristics, we propose smoothing the fitness landscape to facilitate protein optimization. First, we formulate protein fitness as a graph signal then use Tikunov regularization to smooth the fitness landscape. We find optimizing in this smoothed landscape leads to improved performance across multiple methods in the GFP and AAV benchmarks. Second, we achieve state-of-the-art results utilizing discrete energy-based models and MCMC in the smoothed landscape. Our method, called Gibbs sampling with Graph-based Smoothing (GGS), demonstrates a unique ability to achieve 2.5 fold fitness improvement (with in-silico evaluation) over its training set. GGS demonstrates potential to optimize proteins in the limited data regime. Code: https://github.com/kirjner/GGSKeywords
All Related Versions
This publication has 5 references indexed in Scilit:
- Graph Filters for Signal Processing and Machine Learning on GraphsIEEE Transactions on Signal Processing, 2024
- Plug & play directed evolution of proteins with gradient-based discrete MCMCMachine Learning: Science and Technology, 2023
- The adaptive landscape of a metallo-enzyme is shaped by environment-dependent epistasisNature Communications, 2021
- Informed Proposals for Local MCMC in Discrete SpacesJournal of the American Statistical Association, 2019
- An Analysis of NK Landscapes: Interaction Structure, Statistical Properties, and Expected Number of Local OptimaIEEE Transactions on Evolutionary Computation, 2013