Original Research ArticleIntegration of Biomass Formulations of Genome-Scale Metabolic Models with Experimental Data Reveals Universally Essential Cofactors in Prokaryotes

- Seventy one biomass equations of manually curated genome-scale metabolic models are compared.
- Eight classes of universally essential prokaryotic organic cofactors are proposed.
- Conditionally essential organic cofactors are presented and discussed.
- Gene essentiality predictions for Klebsiella pneumoniae are revised.
- A missing essential pathway in models of Mycobacterium tuberculosis is predicted.

The composition of a cell in terms of macromolecular building blocks and other organic molecules underlies the metabolic needs and capabilities of a species. Although some core biomass components such as nucleic acids and proteins are evident for most species, the essentiality of the pool of other organic molecules, especially cofactors and prosthetic groups, is yet unclear. Here we integrate biomass compositions from 71 manually curated genome-scale models, 33 large-scale gene essentiality datasets, enzyme-cofactor association data and a vast array of publications, revealing universally essential cofactors for prokaryotic metabolism and also others that are specific for phylogenetic branches or metabolic modes. Our results revise predictions of essential genes in Klebsiella pneumoniae and identify missing biosynthetic pathways in models of Mycobacterium tuberculosis. This work provides fundamental insights into the essentiality of organic cofactors and has implications for minimal cell studies as well as for modeling genotype-phenotype relations in prokaryotic metabolic networks.