Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Predicting time-varying flux and balance in metabolic systems using structured neural-ODE processes

Santanu Rathod, Pietro Lio, Xiao Zhang

TL;DR

A novel data-driven framework is developed, which trains a structured neural ODE process (SNODEP) model to estimate flux and balance samples using gene-expression time-series data, designed to circumvent the limitations of the standard neural ODE process model.

Abstract

We develop a novel data-driven framework as an alternative to dynamic flux balance analysis, bypassing the demand for deep domain knowledge and manual efforts to formulate the optimization problem. The proposed framework is end-to-end, which trains a structured neural ODE process (SNODEP) model to estimate flux and balance samples using gene-expression time-series data. SNODEP is designed to circumvent the limitations of the standard neural ODE process model, including restricting the latent and decoder sampling distributions to be normal and lacking structure between context points for calculating the latent, thus more suitable for modeling the underlying dynamics of a metabolic system. Through comprehensive experiments ($156$ in total), we demonstrate that SNODEP not only predicts the unseen time points of real-world gene-expression data and the flux and balance estimates well but can even generalize to more challenging unseen knockout configurations and irregular data sampling scenarios, all essential for metabolic pathway analysis. We hope our work can serve as a catalyst for building more scalable and powerful models for genome-scale metabolic analysis. Our code is available at: \url{https://github.com/TrustMLRG/SNODEP}.

Predicting time-varying flux and balance in metabolic systems using structured neural-ODE processes

TL;DR

A novel data-driven framework is developed, which trains a structured neural ODE process (SNODEP) model to estimate flux and balance samples using gene-expression time-series data, designed to circumvent the limitations of the standard neural ODE process model.

Abstract

We develop a novel data-driven framework as an alternative to dynamic flux balance analysis, bypassing the demand for deep domain knowledge and manual efforts to formulate the optimization problem. The proposed framework is end-to-end, which trains a structured neural ODE process (SNODEP) model to estimate flux and balance samples using gene-expression time-series data. SNODEP is designed to circumvent the limitations of the standard neural ODE process model, including restricting the latent and decoder sampling distributions to be normal and lacking structure between context points for calculating the latent, thus more suitable for modeling the underlying dynamics of a metabolic system. Through comprehensive experiments ( in total), we demonstrate that SNODEP not only predicts the unseen time points of real-world gene-expression data and the flux and balance estimates well but can even generalize to more challenging unseen knockout configurations and irregular data sampling scenarios, all essential for metabolic pathway analysis. We hope our work can serve as a catalyst for building more scalable and powerful models for genome-scale metabolic analysis. Our code is available at: \url{https://github.com/TrustMLRG/SNODEP}.

Paper Structure

This paper contains 29 sections, 13 equations, 16 figures, 5 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Problem Formulation
  4. Methodology
  5. Issues with Standard Neural ODE Process
  6. Structured Neural ODE Process (SNODEP)
  7. Optimization objective
  8. Experiments
  9. Experimental Settings
  10. Gene-expression
  11. Metabolic-flux
  12. Metabolic-balance
  13. IRREGULARLY SAMPLED TIMESTEPS
  14. Conclusion and Future Work
  15. Terminology
  16. ...and 14 more sections

Figures (16)

  • Figure 1: Overall pipeline of our framework for predicting time-varying distributions, such as gene expressions, flux, and balance, with (green) and without (orange) gene knockouts.
  • Figure 2: Illustration of the overall pipeline of the proposed SNODEP architecture.
  • Figure 3: Comparison of test-MSE in log-scale between NP, NODEP, and SNODEP across different metabolic pathways on ground-truth gene-expression time-series data.
  • Figure 4: Predicting Metabolic-flux without Gene-knockout
  • Figure 5: Predicting Metabolic-flux with Gene-knockout
  • ...and 11 more figures