CatRange enables robust prediction of enzyme variant kinetic regimes.
6 affiliations
- 1Ames National Laboratory , Ames, IA 50011 ,
- 2Department of Chemical and Biological Engineering, Iowa State University , Ames, IA 50011 ,
- 3Srinivasa Ramanujan Institute for Basic Sciences , Pi Hills, 8th Mile, Velloor, Pampady, Kottayam, Kerala 686501 ,
- 4The Center for Biorenewable Chemicals, Iowa State University , Ames, IA 50011 ,
- 5Department of Chemical and Biomolecular Engineering, University of Nebraska-Lincoln , Lincoln, NE 68588 ,
- 6Department of Computer Science and Engineering, University of South Florida , Tampa, FL 33620 ,
Abstract
Predicting enzyme kinetics directly from sequence remains a central challenge in computational biology, particularly for mutations at catalytically essential residues. Existing models often default to wild-type predictions even when mutations cause substantial activity loss. This limits their utility for enzyme design and mechanistic inference. Here, we introduce CatRange, a machine learning framework trained on CatLog-27k, a human-in-the-loop, AI-agent curated dataset of 27,176 in vitro enzyme-substrate kinetic records. CatLog-27k was created through systematic audit and correction of BRENDA and SABIO-RK, with mutant entries manually reconciled against 2,158 source articles. CatRange reframes kinetic prediction as classification into log₁₀-spaced bins for catalytic turnover (k cat ) and substrate affinity (KM ), rather than exact-value regression. This formulation reflects the order-of-magnitude scale at which enzyme kinetics are commonly interpreted. It reduces sensitivity to assay variability while preserving distinctions among functional catalytic and binding states. Using joint enzyme-substrate representations and gradient-boosted classifiers, CatRange predicts kinetic ranges for wild-type and mutant enzymes across held-out, out-of-distribution, and few-shot mutation settings. The model achieves robust order-of-magnitude recovery and class-balanced discrimination. Moreover, it captures mutation-induced transitions across kinetic regimes, including activity losses caused by perturbations of annotated catalytic residues. CatRange also identifies nonenzyme sequence inputs. Together, CatLog and CatRange provide an interpretable, mutation-sensitive framework for enzyme engineering and kinetic metabolic modeling.
Related
- Conformational Barrier-Driven Flux Redistribution across Competing Catalytic Pathways Unifies Diverse Enzyme Kinetic Regimes
- cPRIME enables endogenous mapping of genetic variant effects at RNA level
- Saturation mutagenesis enables RAD51D clinical variant interpretation and identifies distinct biochemical functions
- Condition aware learning enables robust prediction of oligonucleotide melting behavior across diverse chemistries and assay conditions
- Water as an Active Medium: Distinct Structural-Dynamic Regimes Revealed by Trace Perturbations
BibTeX
@article{sajeevan2026catrange,
title = {{CatRange enables robust prediction of enzyme variant kinetic regimes.}},
author = {Karuna Anna Sajeevan and Abraham Osinuga and B. Arunraj and Sakib Ferdous and Nabia Shahreen and Mohammed Sakib Noor and Shashank Koneru and Laura Mariana Santos-Correa and Rahil Salehi and Niaz Bahar Chowdhury and Randy Aryee and Brisa Calderon-Lopez and Supantha Dey and Ankur Mali and Rajib Saha and Ratul Chowdhury},
journal = {PNAS nexus},
year = {2026},
month = sep,
doi = {10.1093/pnasnexus/pgag309},
url = {https://europepmc.org/article/MED/42872055},
note = {CC BY},
}
RIS
TY - JOUR
TI - CatRange enables robust prediction of enzyme variant kinetic regimes.
AU - Karuna Anna Sajeevan
AU - Abraham Osinuga
AU - B. Arunraj
AU - Sakib Ferdous
AU - Nabia Shahreen
AU - Mohammed Sakib Noor
AU - Shashank Koneru
AU - Laura Mariana Santos-Correa
AU - Rahil Salehi
AU - Niaz Bahar Chowdhury
AU - Randy Aryee
AU - Brisa Calderon-Lopez
AU - Supantha Dey
AU - Ankur Mali
AU - Rajib Saha
AU - Ratul Chowdhury
PY - 2026
DA - 2026/09/16
JO - PNAS nexus
DO - 10.1093/pnasnexus/pgag309
UR - https://europepmc.org/article/MED/42872055
AB - Predicting enzyme kinetics directly from sequence remains a central challenge in computational biology, particularly for mutations at catalytically essential residues. Existing models often default to wild-type predictions even when mutations cause substantial activity loss. This limits their utility for enzyme design and mechanistic inference. Here, we introduce CatRange, a machine learning framework trained on CatLog-27k, a human-in-the-loop, AI-agent curated dataset of 27,176 in vitro enzyme-substrate kinetic records. CatLog-27k was created through systematic audit and correction of BRENDA and SABIO-RK, with mutant entries manually reconciled against 2,158 source articles. CatRange reframes kinetic prediction as classification into log₁₀-spaced bins for catalytic turnover (k cat ) and substrate affinity (KM ), rather than exact-value regression. This formulation reflects the order-of-magnitude scale at which enzyme kinetics are commonly interpreted. It reduces sensitivity to assay variability while preserving distinctions among functional catalytic and binding states. Using joint enzyme-substrate representations and gradient-boosted classifiers, CatRange predicts kinetic ranges for wild-type and mutant enzymes across held-out, out-of-distribution, and few-shot mutation settings. The model achieves robust order-of-magnitude recovery and class-balanced discrimination. Moreover, it captures mutation-induced transitions across kinetic regimes, including activity losses caused by perturbations of annotated catalytic residues. CatRange also identifies nonenzyme sequence inputs. Together, CatLog and CatRange provide an interpretable, mutation-sensitive framework for enzyme engineering and kinetic metabolic modeling.
SN - 2752-6542
ER -