Enabling population protein dynamics through Bayesian modeling

TL;DR

This paper introduces a Bayesian modeling approach to study protein turnover in populations, enabling insights into disease dynamics and biomarker discovery.

Contribution

A novel Bayesian modeling approach is introduced for capturing population-level protein dynamics and inter-individual variability.

Findings

Bayesian models inspired by population pharmacokinetics accurately capture protein turnover in cohorts.

The approach accounts for inter-individual variability, enabling comparative studies of altered dynamics in diseases.

Abstract

The knowledge of protein dynamics, or turnover, in patients provides invaluable information related to certain diseases, drug efficacy, or biological processes. A great corpus of experimental and computational methods has been developed, including by us, in the case of human patients followed in vivo. Moving one step further, we propose a novel modeling approach to capture population protein dynamics using Bayesian methods.

Using two datasets, we demonstrate that models inspired by population pharmacokinetics can accurately capture protein turnover within a cohort and account for inter-individual variability. Such models pave the way for comparative studies searching for altered dynamics or biomarkers in diseases.

R code and preprocessed data are available from zenodo.org. Raw data are available from panoramaweb.org.

Figures