Facilitating Machine Learning‐Guided Protein Engineering with Smart Library Design and Massively Parallel Assays

Abstract

Protein design plays an important role in recent medical advances from antibody therapy to vaccine design. Typically, exhaustive mutational screens or directed evolution experiments are used for the identification of the best design or for improvements to the wild-type variant. Even with a high-throughput screening on pooled libraries and Next-Generation Sequencing to boost the scale of read-outs, surveying all the variants with combinatorial mutations for their empirical fitness scores is still of magnitudes beyond the capacity of existing experimental settings. To tackle this challenge, in-silico approaches using machine learning to predict the fitness of novel variants based on a subset of empirical measurements are now employed. These machine learning models turn out to be useful in many cases, with the premise that the experimentally determined fitness scores and the amino-acid descriptors of the models are informative. The machine learning models can guide the search for the highest fitness variants, resolve complex epistatic relationships, and highlight bio-physical rules for protein folding. Using machine learning-guided approaches, researchers can build more focused libraries, thus relieving themselves from labor-intensive screens and fast-tracking the optimization process. Here, we describe the current advances in massive-scale variant screens, and how machine learning and mutagenesis strategies can be integrated to accelerate protein engineering. More specifically, we examine strategies to make screens more economical, informative, and effective in discovery of useful variants.

State-of-the-art in engineering small molecule biosensors and their applications in metabolic engineering

“…Recent research had utilized machine learning algorithms to predict the optimal design of variant libraries from sequence-function datasets. This approach eliminates the requirement for a profound understanding of deep structure-function relationships, thereby accelerating the process of directed protein evolution [293,294] and the development of biosensors [295] as reviewed by Chu et al in 2021 [296]. Collectively, this research demonstrates the potential of combining these approaches with computational tools to create biosensors with enhanced functionality and specificity.…”

Programmable synthetic receptors: the next-generation of cell and gene therapies

“…In silico approaches enable scientists to construct numerous receptor variants and conduct thousands of simulated experiments by computer, which dramatically reduces the amount of candidates to be tested in laboratory experiments. 222 , 223 …”

Engineering ACE2 decoy receptors to combat viral escapability

“…The other is the machine learning-guided approach. Datasets from DMS and empirical affinity test for multiple mutations are used to train machine learning algorithm, and the resultant machine learning model predicts the affinity value for all possible combination of mutations [ 42 , 43 ].…”