Peter D. Tonner

The paper compares biophysical and phenomenological models to predict epistasis in allosteric proteins, finding biophysical models better capture mutational effects with less epistasis.

StratoMod predicts variant calling errors using interpretable machine learning to help design better variant calling pipelines.

The paper presents two methods for precision engineering of genetic sensors using large-scale data and machine learning to achieve quantitative dose-response specifications.

LANTERN is a hierarchical Bayesian model that improves interpretability and predictive power in genotype-phenotype landscape analysis, outperforming alternatives across benchmark datasets.

The paper predicts human proteins cleaved by SARS-CoV-2 Nsp5 protease using NetCorona and structural analysis, identifying 92 proteins with potential cleavage sites and 48 in the same cellular compartment.

Dose-response curves for 62 472 LacI coding sequences train a recurrent network for EC50 and G∞, enable EC50 targeting within 1.25-fold, and reveal band-stop sensors.

The paper shows that adapting bacterial strains to growth conditions improves the stability of engineered functions across generations without genomic changes.

This paper introduces a Bayesian non-parametric model to analyze microbial growth data, accounting for random effects and improving accuracy in estimating biological effects.