DOI: 10.3390/microorganisms14102187 ISSN: 2076-2607

BEGREM: A Mechanistic Bioelectrochemical Model Coupling Gut Microbial Metabolism

James Torres-Peralta, Josue Duarte-Cuesta, Ronny Amaguay-Gomez, María Román-Manzano, Samuel Valle-Asan

Colonic microbial metabolism emerges from coupled substrate utilization, cross-feeding, redox regulation, extracellular electron transfer (EET), transient reducing-equivalent storage, and spatial transport. We developed BEGREM, a data-constrained mechanistic bioelectrochemical framework that integrates these processes while treating current as a composite observable rather than a direct surrogate of total metabolism. Public data from 22 studies comprised 103,320 current-voltage-power points, 3996 cyclic-voltammetry points, 213 impedance points, and 99 quality-controlled polarization curves. The archived full-BEGREM analysis yielded a median held-out RMSE of 12.4 mV. Independent reconstruction reproduced this scale at 12.147 mV (95% bootstrap: 10.406–14.736 mV) and improved 81/99 curves versus the recoverable Ohmic/Quadratic comparator set, with median ΔRMSE=7.700 mV (95% interval: 4.695–18.175; p=7.20×10−11). Full BEGREM also showed lower AICc in 95/99 curves. OCP-repolarization identified Qmax=3.575 mC and kch=0.02063s−1 with profile-likelihood 95% intervals of 2.302–5.023 mC and 0.00634–0.10633 s−1. Study-held-out transfer produced RMSEs of 75.94 and 104.15 mV, indicating substantial cross-architecture penalties. BEGREM therefore provides a reproducible mechanistic observation framework with strong within-curve and module-level support, while colon-specific calibration remains necessary.