Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Integrating Optimal Transport and Structural Inference Models for GRN Inference from Single-cell Data

Tsz Pan Tong, Aoran Wang, George Panagopoulos, Jun Pang

TL;DR

A novel gene regulatory network (GRN) inference method that integrates optimal transport (OT) with a deep-learning structural inference model that recovers the GRN from single-cell data is introduced.

Abstract

We introduce a novel gene regulatory network (GRN) inference method that integrates optimal transport (OT) with a deep-learning structural inference model. Advances in next-generation sequencing enable detailed yet destructive gene expression assays at the single-cell level, resulting in the loss of cell evolutionary trajectories. Due to technological and cost constraints, single-cell experiments often feature cells sampled at irregular and sparse time points with a small sample size. Although trajectory-based structural inference models can accurately reveal the underlying interaction graph from observed data, their efficacy depends on the inputs of thousands of regularly sampled trajectories. The irregularly-sampled nature of single-cell data precludes the direct use of these powerful models for reconstructing GRNs. Optimal transport, a classical mathematical framework that minimize transportation costs between distributions, has shown promise in multi-omics data integration and cell fate prediction. Utilizing OT, our method constructs mappings between consecutively sampled cells to form cell-level trajectories, which are given as input to a structural inference model that recovers the GRN from single-cell data. Through case studies in two synthetic datasets, we demonstrate the feasibility of our proposed method and its promising performance over eight state-of-the-art GRN inference methods.

Integrating Optimal Transport and Structural Inference Models for GRN Inference from Single-cell Data

TL;DR

A novel gene regulatory network (GRN) inference method that integrates optimal transport (OT) with a deep-learning structural inference model that recovers the GRN from single-cell data is introduced.

Abstract

We introduce a novel gene regulatory network (GRN) inference method that integrates optimal transport (OT) with a deep-learning structural inference model. Advances in next-generation sequencing enable detailed yet destructive gene expression assays at the single-cell level, resulting in the loss of cell evolutionary trajectories. Due to technological and cost constraints, single-cell experiments often feature cells sampled at irregular and sparse time points with a small sample size. Although trajectory-based structural inference models can accurately reveal the underlying interaction graph from observed data, their efficacy depends on the inputs of thousands of regularly sampled trajectories. The irregularly-sampled nature of single-cell data precludes the direct use of these powerful models for reconstructing GRNs. Optimal transport, a classical mathematical framework that minimize transportation costs between distributions, has shown promise in multi-omics data integration and cell fate prediction. Utilizing OT, our method constructs mappings between consecutively sampled cells to form cell-level trajectories, which are given as input to a structural inference model that recovers the GRN from single-cell data. Through case studies in two synthetic datasets, we demonstrate the feasibility of our proposed method and its promising performance over eight state-of-the-art GRN inference methods.
Paper Structure (18 sections, 4 equations, 2 figures, 2 tables)

This paper contains 18 sections, 4 equations, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Gene Regulatory Networks Inference
  4. Structural Inference
  5. Preliminaries
  6. Notations and Problem Definition
  7. Waddington-OT (WOT)
  8. Neural Relational Inference (NRI)
  9. Our Method
  10. Experiments
  11. Experiment Setup
  12. Evaluation on Trajectory Reconstruction
  13. Evaluation Metrics on Structural Inference
  14. Results
  15. Discussion
  16. ...and 3 more sections

Figures (2)

  • Figure 1: Overview of the proposed pipeline. With gene expression matrixes at multiple time points as input, we first connect cells at consecutive time points using WOT. The reconstructed cell evolutionary trajectories are then fed into the NRI model. The inferred GRN is the output of the trained NRI encoder.
  • Figure 2: Comparison of normalized trajectory reconstruction error from time $i$ to $i+1$ in different GRNs between WOT, OT and random.