Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Interpretable deep learning in single-cell omics

Manoj M Wagle, Siqu Long, Carissa Chen, Chunlei Liu, Pengyi Yang

TL;DR

The basics of single-cell omics technologies and the concept of interpretable deep learning are introduced and a review of the recent interpretable deep learning models applied to various single-cell omics research is reviewed.

Abstract

Recent developments in single-cell omics technologies have enabled the quantification of molecular profiles in individual cells at an unparalleled resolution. Deep learning, a rapidly evolving sub-field of machine learning, has instilled a significant interest in single-cell omics research due to its remarkable success in analysing heterogeneous high-dimensional single-cell omics data. Nevertheless, the inherent multi-layer nonlinear architecture of deep learning models often makes them `black boxes' as the reasoning behind predictions is often unknown and not transparent to the user. This has stimulated an increasing body of research for addressing the lack of interpretability in deep learning models, especially in single-cell omics data analyses, where the identification and understanding of molecular regulators are crucial for interpreting model predictions and directing downstream experimental validations. In this work, we introduce the basics of single-cell omics technologies and the concept of interpretable deep learning. This is followed by a review of the recent interpretable deep learning models applied to various single-cell omics research. Lastly, we highlight the current limitations and discuss potential future directions. We anticipate this review to bring together the single-cell and machine learning research communities to foster future development and application of interpretable deep learning in single-cell omics research.

Interpretable deep learning in single-cell omics

TL;DR

The basics of single-cell omics technologies and the concept of interpretable deep learning are introduced and a review of the recent interpretable deep learning models applied to various single-cell omics research is reviewed.

Abstract

Recent developments in single-cell omics technologies have enabled the quantification of molecular profiles in individual cells at an unparalleled resolution. Deep learning, a rapidly evolving sub-field of machine learning, has instilled a significant interest in single-cell omics research due to its remarkable success in analysing heterogeneous high-dimensional single-cell omics data. Nevertheless, the inherent multi-layer nonlinear architecture of deep learning models often makes them `black boxes' as the reasoning behind predictions is often unknown and not transparent to the user. This has stimulated an increasing body of research for addressing the lack of interpretability in deep learning models, especially in single-cell omics data analyses, where the identification and understanding of molecular regulators are crucial for interpreting model predictions and directing downstream experimental validations. In this work, we introduce the basics of single-cell omics technologies and the concept of interpretable deep learning. This is followed by a review of the recent interpretable deep learning models applied to various single-cell omics research. Lastly, we highlight the current limitations and discuss potential future directions. We anticipate this review to bring together the single-cell and machine learning research communities to foster future development and application of interpretable deep learning in single-cell omics research.
Paper Structure (17 sections, 2 figures)

This paper contains 17 sections, 2 figures.

Table of Contents

  1. Abstract
  2. Introduction
  3. Fundamentals of single-cell omics and interpretable deep learning
  4. Harnessing the power of interpretable deep learning for single-cell omics research
  5. Current challenges and future opportunities
  6. Conclusions

Figures (2)

  • Figure 1: Summary illustration of single-cell omics. (a) A schematic of single cells from different complex tissues/organs. (b) Molecular attributes in a single cell and their corresponding modalities in single-cell omics. (c, d) Example unimodal (c) and multimodal (d) single-cell omics technologies.
  • Figure 2: Schematic of example interpretable deep learning models applied to single-cell omics. (a) Using post-hoc feature attribution techniques to identify cell identity genes that distinguish cell types huang2023evaluation. Colours denote the estimated importance of input features (b) Designing an intrinsic and model-specific attention layer to detect gene sets for annotating cell types chen2023transformer. Colours denote the estimated importance of latent features (c) Incorporating prior knowledge to design embeddings for detecting activated gene programs underlie cell types lotfollahi2023biologically. Colours denote different gene programs. (d) Using prior knowledge to design neural network architectures for modelling molecular networks fortelny2020knowledge. Colours denote different molecular species.