ML4SD: Leveraging Machine Learning and High-Throughput Search Algorithms for an Iterative Growth-Coupled Design Innovation
Optimizing microbial biomanufacturing is required if renewable and waste carbon are to replace petrochemical routes at competitive titers, rates, and yields. Growth-coupled (GC) production supports that goal by linking target synthesis to biomass formation, so product formation is required for growth. Constructing knockout strains yielding GC production from a list of candidate genes is labor and…
ML4SD is a new tool that uses machine learning and high-throughput search algorithms to optimize microbial biomanufacturing. It focuses on growth-coupled production, where target synthesis is linked to biomass formation. Traditional methods for creating knockout strains to achieve GC production are time-consuming and limit the number of in vivo tested designs.
ML4SD overcomes this by developing an active-learning Design-Build-Test-Learn (DBTL) cycle, which trains ensembles on genome-scale metabolic model (GEM) scores of knockout designs and samples the next designs from predicted model performance and error.
For ML4SD to work effectively, the initial library of designs must be large and diverse, including suboptimal and non-viable options. Libraries that are restricted to minimal designs or Pareto-optimal knockouts can lead to overfitting. To address this, the gcSwarms strain design algorithm was developed and tested on various bioprocesses within Pseudomonas putida iJN1462. In an in silico case study, ML4SD was used to convert lignin-derived 4-hydroxybenzoate to 6-caprolactam, the nylon-6 monomer.
The results showed that ML4SD improved carbon yield by up to 164%, and it achieved this using 2.5- to 7.1-fold fewer designs than a gcSwarms-only search. This demonstrates the data efficiency of ML4SD, which can learn GC patterns for specific bioprocesses with fewer designs.
Written by urgent.news from bioRxiv's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
