Machine Learning Optimization of Candidate Antibodies Yields Highly Diverse Sub-nanomolar Affinity Antibody Libraries

Abstract

Therapeutic antibodies are an important and rapidly growing drug modality. However, the design and discovery of early-stage antibody therapeutics remain a time and cost-intensive endeavor. In this work, we present an end-to-end Bayesian, language model-based method for designing large and diverse libraries of high-affinity single-chain variable fragments (scFvs). We integrate target-specific binding affinities with information from millions of natural protein sequences in a probabilistic machine learning framework to design thousands of scFvs that are then empirically measured. In a head-to-head comparison with a directed evolution approach, we show that the best scFv generated from our method represents a 28.8-fold improvement in binding over the best scFv from the directed evolution. Additionally, 99% of the designed scFvs in our most successful library are improvements over the initial candidate scFv. By comparing a library's predicted success to actual measurements, we demonstrate our method's ability to explore tradeoffs between library success and diversity during the design phase and prior to experimental testing. The results of our work highlight the significant impact machine learning models can have on scFv development. We expect our end-to-end method to be broadly applicable and able to provide value to other protein engineering tasks.

Likelihood-based fine-tuning of protein language models for few-shot fitness prediction and design

“…An alternative in silico evaluation strategy avoids the challenge of defining a meaningful oracle function by adopting a pool-based optimisation problem formulation over experimentally determined fitness landscapes [30]. Another line of works has sought to provide direct experimental validation of approaches combining uncertainty estimates with PLMs, in settings ranging from zero [14] or few-shot design [2] to single-round design given large training sets of sequence-fitness pairs [24]. In this paper, we focus on evaluating different PLM-based fitness prediction strategies in in silico settings designed to mimic applied design scenarios.…”

Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries

“…Code 48 is available on github: https://github.com/AIforGreatGood/ biotransfer and on Zenodo under https://doi.org/10.5281/zenodo. 7927152 for academic and/or non-profit internal research purposes.…”

Baselining the Buzz Trastuzumab-HER2 Affinity, and Beyond

“…Li et al [15] used a pre-trained LLM, Gaussian Processes (GPs), and ensemble regression models to design and screen new high-affinity single-chain fragment variable antibodies (scFvs) against a conserved coronavirus peptide. They trained their models on experimental data (26.5k heavy and 26.2k light chain sequences) that involved up to three random CDR mutations from an initial candidate.…”