Resolving Single-Cell Gene Expression by Pseudotemporal Integration of Transcriptomic and Proteomic Datasets

TL;DR

This study introduces a method to combine single-cell RNA and protein data to better understand how cells respond to low oxygen conditions.

Contribution

A novel pseudotemporal integration method for aligning transcriptomic and proteomic data in hypoxia.

Findings

Transcriptomic responses to hypoxia occur before proteomic changes.

kNN imputation improves detection of hypoxia signals in proteomic data.

Pseudotemporal ordering aligns scRNA-Seq and scp-MS data for multiomic profiling.

Abstract

Single-cell omics technologies, such as single-cell RNA-Seq and single-cell proteomics, offer unprecedented insights into cellular heterogeneity and dynamic regulatory processes. However, integrating these data types to construct comprehensive transcription–translation profiles remains challenging because of their distinct and complex behaviors. This study presents a novel approach using pseudotemporal cell ordering to integrate single-cell RNA-Seq and single-cell proteomics by mass spectrometry data, facilitating the analysis of transcription–translation expression dynamics. We collected longitudinal single-cell samples following hypoxia. By leveraging key markers, we constructed pseudotemporal trajectories for each data type, revealing transcriptional and translational responses to hypoxia. This profile of unified single-cell mRNA and protein expression uncovers distinct regulatory mechanisms, including an immediate transcriptomic response, followed by delayed proteomic expression. It illustrates the use of pseudotemporal integration to integrate single-cell transcriptomic and proteomic datasets to understand the cellular phenotypes under hypoxic stress and provides a framework for future investigations into transcription–translation dynamics.

•Hypoxia progression is defined by distinct transcriptomic and proteomic markers.•Pseudotemporal ordering aligns hypoxia progression in scRNA-Seq and scp-MS data.•kNN imputation enhances detection of hypoxia signals in scp-MS datasets.•Pseudotemporal integration enables multiomic profiling of hypoxic response.

Hypoxia progression is defined by distinct transcriptomic and proteomic markers.

Pseudotemporal ordering aligns hypoxia progression in scRNA-Seq and scp-MS data.

kNN imputation enhances detection of hypoxia signals in scp-MS datasets.

Pseudotemporal integration enables multiomic profiling of hypoxic response.

This study presents a method for integrating unpaired single-cell transcriptomic and proteomic data using pseudotemporal alignment. In HEK 293 cells under hypoxia, the authors show that distinct markers in each modality represent hypoxia progression. k-Nearest neighbor imputation recovers lowly represented hypoxia signals in proteomic data, enabling pseudotemporal ordering to align transcriptomic and proteomic responses. This approach supports joint analysis of gene and protein expression dynamics across the hypoxic trajectory.

Figures