Single-cell RNA-sequencing (scRNA-seq) technologies offer a chance to understand the regulatory mechanisms at single-cell resolution. Gene regulatory networks (GRNs) constitute a crucial blueprint of regulatory mechanisms in cellular systems, thereby playing a pivotal role in biological research. Therefore, it’s imperative to develop a precise tool for inferring GRNs from scRNA-seq data.

Researchers from Wuhan Botanical Garden developed a novel method, namely STGRNS, for constructing GRNs based on scRNA-seq data using deep learning model. The tool and tutorial are publicly available at https://github.com/zhanglab-wbgcas/STGRNS .

In this algorithm, a gene expression motif (GEM) technique is proposed to convert each gene pair into a form that can be received as the transformer encoder. By avoiding missing phase-specific regulations in a network, STGRNS can accurately infer GRNs from static, pseudo-time, or time series single-cell transcriptome data.

This research shows that STGRNS is superior to other state-of-the-art methods as a deep learning-based method on 48 benchmark datasets including 21 static scRNA-seq dataset and 27 time-series scRNA-seq datasets.

Unlike other “black box” deep learning-based methods, which are often characterized by their opacity and associated difficulty in furnishing lucid justifications for their predictions, STGRNS is more reliable and can interpret the predictions. In addition, STGRNS has fewer hyperparameters compared to other GRN reconstruction methods based on deep learning models, which is one of the main reasons for its outstanding generalization.

This research was funded by the National Natural Science Foundation of China. The results have been published in Bioinformatics entitled “STGRNS: an interpretable transformer-based method for inferring gene regulatory networks from single-cell transcriptomic data”. (Available online 2 April, 2023)