Generative machine learning produces kinetic models that accurately characterize intracellular metabolic states
Problem
Determining kinetic parameters for large-scale metabolic models is difficult, hindering integration of omics data and broader adoption of these models.
Why it matters
Kinetic models explicitly connect metabolites, fluxes and enzyme levels, but parameterization bottlenecks block their use.
What the AI did
RENAISSANCE, a generative machine learning framework, parameterizes large-scale kinetic models with dynamic properties matching experimental observations, integrating diverse omics data.
Results
Produced kinetic models with dynamic properties matching experimental observations; no quantitative metrics stated.
Impact
Framework addresses the parameterization bottleneck for kinetic models, per the authors' claims.
Record
| Venue | Nature Catalysis |
|---|---|
| Published | 2024 |
| Work type | article |
| Authors | Subham Choudhury, Bharath Narayanan, Michaël Moret, Vassily Hatzimanikatis, Ljubiša Mišković |
| Citations | 64 (OpenAlex, at import time) |
| References | 70 |
| Model or method | RENAISSANCE |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41929-024-01220-6 |
| Open access | hybrid |
| Keywords | kinetic models, generative machine learning, metabolism, omics data, RENAISSANCE |
Importance rating
real but narrower advance
Addresses kinetic-parameter bottleneck blocking adoption of metabolic models within the research community; demonstrated on E. coli network, a real but narrow methodological advance.