High-throughput generation and comparison of genome-scale metabolic models reveal strain-specific metabolic diversity in 439 Lactococcus strains
Model predictions of growth without different amino acids, and the effects of adding cysteine to casein. Taken from the original publication:
https://doi.org/10.1128/msystems.01517-25Abstract
When developing bacterial starters for the food industry, strain diversity is essential, due to the limited genetic manipulation that can be done. Properly reflecting that diversity in computational simulations is paramount: that is why we developed a method for streamlining this process for Lactococcus. This included adapting an existing method (known as carveme), heavily curating the knowledgebase of bacterial metabolism (known as bacterial-universe), and generating genome-scale models for all sequenced Lactococcus strains in NCBI. The models revealed a lot of interesting things about the diversity in the Lactococcus genus, including different growth behavior under a number of substrates. This project was driven by Jildau and counted with my co-supervision: It was a delight to work with her, and it was very fulfilling to see the work finally published after she successfully completed her PhD. I hope the approach is used by more in the field, and that both the curated universe and the Lactococcus models are useful as a knowledge resource.
