Radhakrishnan Mahadevan

Do protein representations encode variant function? Diagnosing residue-level protein attributions via embedding-space gradients

This paper evaluates residue-level attributions in protein language models using synthetic fitness landscapes and identifies interventions that improve attribution accuracy.

kinGEMs: A Robust and Scalable Framework for Resource-Constraint Models through Stochastic Tuning of Deep Learning-Predicted Kinetic Parameters

This paper introduces kinGEMs, a framework that integrates deep learning-predicted kinetic parameters into genome-scale models to improve metabolic modeling accuracy and scalability.

Cell Geometry and Membrane Protein Crowding Constrain Growth Rate, Overflow Metabolism, Respiration, and Maintenance Energy.

The paper develops a theory linking cell geometry, membrane protein crowding, and metabolic traits in E. coli, validated through experimental data.

Generalist method to reconstruct metabolic networks from multi-omics data at large-scale

The paper introduces SPECTRA, a method to reconstruct metabolic networks from multi-omics data at large-scale, improving speed and reaction inclusion across cancer cell lines, microbial reconstructions, and synthetic gut microbiota.

Enabling Plasmid-based Expression in Clostridium kluyveri using a Biparental Methylation-Conjugation System

This paper develops a methylation-conjugation system for Clostridium kluyveri to enable plasmid-based genetic modification and demonstrates its use for expressing a fluorescent reporter.

Synthetic metabolic pathways for ethylene glycol assimilation outperform natural counterparts

This study compares synthetic and natural metabolic pathways for ethylene glycol assimilation, finding synthetic pathways offer higher thermodynamic driving forces and yields but natural pathways are more orthogonal to central metabolism.

Modeling reveals metabolic basis of competition among Dehalobacter strains during tandem CF and DCM metabolism

The paper models metabolic differences between two Dehalobacter strains to show how their genetic makeup affects their competitive advantage in chloroform and dichloromethane metabolism.

CPI-Pred: A deep learning framework for predicting functional parameters of compound-protein interactions

CPI-Pred predicts compound-protein interaction parameters using deep learning, outperforming existing models on unseen compounds and enzymes.

Functional and Structural Characterization of an IclR Family Transcription Factor for the Development of Dicarboxylic Acid Biosensors

This paper characterizes the PcaR transcription factor, solving its crystal structures and demonstrating its function as a dicarboxylic acid biosensor in yeast.

Understanding systems level metabolic adaptation resulting from osmotic stress

This study develops a novel modeling approach to analyze metabolic adaptation in E. coli under osmotic stress, revealing prioritization of thermodynamically favorable pathways.

Optimizing Fed-Batch Processes with Dynamic Control Flux Balance Analysis

This paper improves dynamic flux balance analysis for fed-batch processes by reducing KKT constraints and adapting the algorithm to predict optimal process length and feeding profiles.

Evaluating the feasibility of medium chain oleochemical synthesis using microbial chain elongation

This study evaluates the feasibility of expanding the product range of chain elongating bacteria to produce C9-C12 fatty acids and other derivatives using metabolic engineering and constraint-based modelling.

Standard-GEM supports reproducible development of genome-scale metabolic models

Standard-GEM provides a template for reproducible genome-scale metabolic model development following FAIR principles and software best practices.

Increasing evenness and stability in synthetic microbial consortia

This study investigates how social interactions in synthetic microbial consortia affect evenness and stability, providing insights for optimizing community design.

Design principles for bistable genetic switches emerge from a growth-dependent speed-robustness trade-off

The paper identifies design principles for bi-stable genetic switches by analyzing the trade-off between switching speed and robustness, validated experimentally in E. coli.

A gap-filling algorithm for prediction of metabolic interactions in microbial communities

The paper introduces a community-level gap-filling algorithm to resolve metabolic gaps and predict interactions in microbial communities.

SARS-CoV-2 receptor binding mutations and antibody mediated immunity.

The paper analyzes SARS-CoV-2 receptor binding domain mutations and their impact on antibody binding using GISAID sequences and PDBsum data.

Thermodynamic constraints are sufficient for the emergence of flux sensors in metabolism

The paper shows that substrate concentrations in thermodynamically constrained reactions reflect upstream fluxes and can transmit information about metabolic rates in E. coli.

Novel two-stage processes for optimal chemical production in microbes

This paper develops a computational framework to identify optimal growth and production phenotypes for two-stage microbial processes, showing that intermediate growth during production leads to better TRY metrics.

Pan-genome-scale network reconstruction: a framework to increase the quantity and quality of metabolic network reconstructions throughout the tree of life

The paper introduces a pan-genome framework to improve the quantity and quality of metabolic network reconstructions across fungal species.

Memote: A community-driven effort towards a standardized genome-scale metabolic model test suite

Memote, open-source Python software, runs a community-maintained, partly scored test suite (consistency, biomass, annotations, optional experimental data) on SBML-encoded genome-scale metabolic models, once or through continuous integration.

AYbRAH: a curated ortholog database for yeasts and fungi spanning 600 million years of evolution

AYbRAH is a curated ortholog database for yeasts and fungi, comparing ortholog groups across multiple phylogenomic databases to identify discrepancies and improve accuracy.

A Multi-Scale Modelling Framework of the Whole Human Body Based on Multi-Objective Functions

This paper develops a multi-scale metabolic model of the human body with 3555 ODEs and 1826 reactions, solving 74 objectives to simulate metabolic phenotypes.

Flux balance analysis predicts NADP phosphatase and NADH kinase are critical to balancing redox during xylose fermentation in Scheffersomyces stipitis

Genome-scale metabolic network reconstruction of the chloroform-respiring Dehalobacter restrictus strain CF

The paper reconstructs and curates a genome-scale metabolic network for Dehalobacter restrictus strain CF, highlighting key metabolic pathways and the need for manual curation in automated reconstructions.