Evolutionary Sequential Monte Carlo Samplers for Change-Point Models

Dufays, Arnaud

doi:10.3390/econometrics4010012

Cited by 9 publications

(1 citation statement)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Evolutionary Monte Carlo (EMC) [26,27] is an advanced sampling method that draws inspiration from genetic recombination as well as physics-based MCMC techniques. While EMC has previously been used for a variety of sampling tasks [26,[28][29][30][31], its potential as an exploratory algorithm for protein design remains unexplored. In this paper, we modify EMC as a search tool for exploring the complex fitness landscape of protein sequences capable of gene regulation, which we call EMC Search (EMCS).…”

Section: Introductionmentioning

confidence: 99%

Improving few-shot learning-based protein engineering with evolutionary sampling

Jawaid¹,

Yeo²,

Gautam³

et al. 2023

Preprint

View full text Add to dashboard Cite

Designing novel functional proteins remains a slow and expensive process due to a variety of protein engineering challenges; in particular, the number of protein variants that can be experimentally tested in a given assay pales in comparison to the vastness of the overall sequence space, resulting in low hit rates and expensive wet lab testing cycles. In this paper, we propose a few-shot learning approach to novel protein design that aims to accelerate the expensive wet lab testing cycle and is capable of leveraging a training dataset that is both small and skewed (≈ 10^5 datapoints, < 1% positive hits). Our approach is composed of two parts: a semi-supervised transfer learning approach to generate a discrete fitness landscape for a desired protein function and a novel evolutionary Monte Carlo Markov Chain sampling algorithm to more efficiently explore the fitness landscape. We demonstrate the performance of our approach by experimentally screening predicted high fitness gene activators, resulting in a dramatically improved hit rate compared to existing methods. Our method can be easily adapted to other protein engineering and design problems, particularly where the cost associated with obtaining labeled data is significantly high. We have provided open source code for our method at github.com/SuperSecretBioTech/evolutionary_monte_carlo_search.

show abstract