TENET: gene network reconstruction using transfer entropy reveals key regulatory factors from single cell transcriptomic data

Topological benchmarking of algorithms to infer gene regulatory networks from single-cell RNA-seq data

MOTIVATION

In recent years, many algorithms for inferring gene regulatory networks from single-cell transcriptomic data have been published. Several studies have evaluated their accuracy in estimating the presence of an interaction between pairs of genes. However, these benchmarking analyses do not quantify the algorithms' ability to capture structural properties of networks, which are fundamental, e.g., for studying the robustness of a gene network to external perturbations. Here, we devise a three-step benchmarking pipeline called STREAMLINE that quantifies the ability of algorithms to capture topological properties of networks and identify hubs.

RESULTS

To this aim, we use data simulated from different types of networks as well as experimental data from three different organisms. We apply our benchmarking pipeline to four inference algorithms and provide guidance on which algorithm should be used depending on the global network property of interest.

AVAILABILITY AND IMPLEMENTATION

STREAMLINE is available at https://github.com/ScialdoneLab/STREAMLINE. The data generated in this study are available at https://doi.org/10.5281/zenodo.10710444.

Prediction of protein-RNA interactions from single-cell transcriptomic data

Proteins are crucial in regulating every aspect of RNA life, yet understanding their interactions with coding and noncoding RNAs remains limited. Experimental studies are typically restricted to a small number of cell lines and a limited set of RNA-binding proteins (RBPs). Although computational methods based on physico-chemical principles can predict protein-RNA interactions accurately, they often lack the ability to consider cell-type-specific gene expression and the broader context of gene regulatory networks (GRNs). Here, we assess the performance of several GRN inference algorithms in predicting protein-RNA interactions from single-cell transcriptomic data, and propose a pipeline, called scRAPID (single-cell transcriptomic-based RnA Protein Interaction Detection), that integrates these methods with the catRAPID algorithm, which can identify direct physical interactions between RBPs and RNA molecules. Our approach demonstrates that RBP-RNA interactions can be predicted from single-cell transcriptomic data, with performances comparable or superior to those achieved for the well-established task of inferring transcription factor-target interactions. The incorporation of catRAPID significantly enhances the accuracy of identifying interactions, particularly with long noncoding RNAs, and enables the identification of hub RBPs and RNAs. Additionally, we show that interactions between RBPs can be detected based on their inferred RNA targets. The software is freely available at https://github.com/tartaglialabIIT/scRAPID.

TFvelo: gene regulation inspired RNA velocity estimation

RNA velocity is closely related with cell fate and is an important indicator for the prediction of cell states with elegant physical explanation derived from single-cell RNA-seq data. Most existing RNA velocity models aim to extract dynamics from the phase delay between unspliced and spliced mRNA for each individual gene. However, unspliced/spliced mRNA abundance may not provide sufficient signal for dynamic modeling, leading to poor fit in phase portraits. Motivated by the idea that RNA velocity could be driven by the transcriptional regulation, we propose TFvelo, which expands RNA velocity concept to various single-cell datasets without relying on splicing information, by introducing gene regulatory information. Our experiments on synthetic data and multiple scRNA-Seq datasets show that TFvelo can accurately fit genes dynamics on phase portraits, and effectively infer cell pseudo-time and trajectory from RNA abundance data. TFvelo opens a robust and accurate avenue for modeling RNA velocity for single cell data.

MICA: a multi-omics method to predict gene regulatory networks in early human embryos

Recent advances in single-cell omics have transformed characterisation of cell types in challenging-to-study biological contexts. In contexts with limited single-cell samples, such as the early human embryo inference of transcription factor-gene regulatory network (GRN) interactions is especially difficult. Here, we assessed application of different linear or non-linear GRN predictions to single-cell simulated and human embryo transcriptome datasets. We also compared how expression normalisation impacts on GRN predictions, finding that transcripts per million reads outperformed alternative methods. GRN inferences were more reproducible using a non-linear method based on mutual information (MI) applied to single-cell transcriptome datasets refined with chromatin accessibility (CA) (called MICA), compared with alternative network prediction methods tested. MICA captures complex non-monotonic dependencies and feedback loops. Using MICA, we generated the first GRN inferences in early human development. MICA predicted co-localisation of the AP-1 transcription factor subunit proto-oncogene JUND and the TFAP2C transcription factor AP-2γ in early human embryos. Overall, our comparative analysis of GRN prediction methods defines a pipeline that can be applied to single-cell multi-omics datasets in especially challenging contexts to infer interactions between transcription factor expression and target gene regulation.

A review on gene regulatory network reconstruction algorithms based on single cell RNA sequencing

BACKGROUND

Understanding gene regulatory networks (GRNs) is essential for unraveling the molecular mechanisms governing cellular behavior. With the advent of high-throughput transcriptome measurement technology, researchers have aimed to reverse engineer the biological systems, extracting gene regulatory rules from their outputs, which represented by gene expression data. Bulk RNA sequencing, a widely used method for measuring gene expression, has been employed for GRN reconstruction. However, it falls short in capturing dynamic changes in gene expression at the level of individual cells since it averages gene expression across mixed cell populations.

OBJECTIVE

In this review, we provide an overview of 15 GRN reconstruction tools and discuss their respective strengths and limitations, particularly in the context of single cell RNA sequencing (scRNA-seq).

METHODS

Recent advancements in scRNA-seq break new ground of GRN reconstruction. They offer snapshots of the individual cell transcriptomes and capturing dynamic changes. We emphasize how these technological breakthroughs have enhanced GRN reconstruction.

CONCLUSION

GRN reconstructors can be classified based on their requirement for cellular trajectory, which represents a dynamical cellular process including differentiation, aging, or disease progression. Benchmarking studies support the superiority of GRN reconstructors that do not require trajectory analysis in identifying regulator-target relationships. However, methods equipped with trajectory analysis demonstrate better performance in identifying key regulatory factors. In conclusion, researchers should select a suitable GRN reconstructor based on their specific research objectives.

Deciphering driver regulators of cell fate decisions from single-cell transcriptomics data with CEFCON

Single-cell technologies enable the dynamic analyses of cell fate mapping. However, capturing the gene regulatory relationships and identifying the driver factors that control cell fate decisions are still challenging. We present CEFCON, a network-based framework that first uses a graph neural network with attention mechanism to infer a cell-lineage-specific gene regulatory network (GRN) from single-cell RNA-sequencing data, and then models cell fate dynamics through network control theory to identify driver regulators and the associated gene modules, revealing their critical biological processes related to cell states. Extensive benchmarking tests consistently demonstrated the superiority of CEFCON in GRN construction, driver regulator identification, and gene module identification over baseline methods. When applied to the mouse hematopoietic stem cell differentiation data, CEFCON successfully identified driver regulators for three developmental lineages, which offered useful insights into their differentiation from a network control perspective. Overall, CEFCON provides a valuable tool for studying the underlying mechanisms of cell fate decisions from single-cell RNA-seq data.

Network Analysis and Experimental Verification of the Mechanisms of Hydroxysafflor Yellow A in Ischemic Stroke Following Atherosclerosis

Hydroxysafflor yellow A (HSYA) is derived from Carthamus tinctorius L. (Honghua in Chinese) and is used to treat cardiovascular and cerebrovascular disease. However, the mechanism by which HSYA treats ischemic stroke following atherosclerosis (ISFA) remains unclear. The targets and pathways of HSYA against ISFA were obtained using network analysis. A total of 3335 potential IFSA-related targets were predicted using the GenCards and Drugbank databases, and a total of 88 potential HSYA-related targets were predicted using the Swiss Target Prediction database. A total of 62 HSYA-related targets against IFSA were obtained. The network was composed of HSYA, 62 targets, and 20 pathways. The top 20 targets were constructed via the protein-protein interaction (PPI) network. Gene Ontology analysis revealed that the targets were involved in signal transduction, protein phosphorylation, the cytoplasm, the plasma membrane, the cytosol, zinc ion binding, ATP binding, protein kinase binding/activity, and enzyme binding. The Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis revealed that the pathways were associated with cancer, inflammatory mediator regulation of the transient receptor potential channels, and microRNA in cancer. Additionally, molecular docking indicated that HSYA mainly interacts with five targets, namely interleukin 1 beta (IL-1β), signal transducer and activator of transcription 3 (STAT3), E1A-binding protein p300 (EP300), protein kinase C alpha (PRKCA), and inhibitor of nuclear factor kappa B kinase subunit beta (IKBKB). In animal experiments, HSYA administration ameliorated the infarct size, neurological deficit score, histopathological changes, carotid intima-media thickness (IMT), and blood lipid level (total cholesterol and triglycerides). Immunochemistry and quantitative PCR showed that HSYA intervention downregulated the expression of STAT3, EP300, PRKCA, and IKBKB, and the enzyme-linked immunoassay showed reduced IL-1β levels. The findings of this study provide a reference for the development of anti-ISFA drugs.

Exploration of cell state heterogeneity using single-cell proteomics through sensitivity-tailored data-independent acquisition

Single-cell resolution analysis of complex biological tissues is fundamental to capture cell-state heterogeneity and distinct cellular signaling patterns that remain obscured with population-based techniques. The limited amount of material encapsulated in a single cell however, raises significant technical challenges to molecular profiling. Due to extensive optimization efforts, single-cell proteomics by Mass Spectrometry (scp-MS) has emerged as a powerful tool to facilitate proteome profiling from ultra-low amounts of input, although further development is needed to realize its full potential. To this end, we carry out comprehensive analysis of orbitrap-based data-independent acquisition (DIA) for limited material proteomics. Notably, we find a fundamental difference between optimal DIA methods for high- and low-load samples. We further improve our low-input DIA method by relying on high-resolution MS1 quantification, thus enhancing sensitivity by more efficiently utilizing available mass analyzer time. With our ultra-low input tailored DIA method, we are able to accommodate long injection times and high resolution, while keeping the scan cycle time low enough to ensure robust quantification. Finally, we demonstrate the capability of our approach by profiling mouse embryonic stem cell culture conditions, showcasing heterogeneity in global proteomes and highlighting distinct differences in key metabolic enzyme expression in distinct cell subclusters.

Gene regulatory network inference with popInfer reveals dynamic regulation of hematopoietic stem cell quiescence upon diet restriction and aging

Inference of gene regulatory networks (GRNs) can reveal cell state transitions from single-cell genomics data. However, obstacles to temporal inference from snapshot data are difficult to overcome. Single-nuclei multiomics data offer means to bridge this gap and derive temporal information from snapshot data using joint measurements of gene expression and chromatin accessibility in the same single cells. We developed popInfer to infer networks that characterize lineage-specific dynamic cell state transitions from joint gene expression and chromatin accessibility data. Benchmarking against alternative methods for GRN inference, we showed that popInfer achieves higher accuracy in the GRNs inferred. popInfer was applied to study single-cell multiomics data characterizing hematopoietic stem cells (HSCs) and the transition from HSC to a multipotent progenitor cell state during murine hematopoiesis across age and dietary conditions. From networks predicted by popInfer, we discovered gene interactions controlling entry to/exit from HSC quiescence that are perturbed in response to diet or aging.

Mathematical Modeling of Eicosanoid Metabolism in Macrophage Cells: Cybernetic Framework Combined with Novel Information-Theoretic Approaches

Cellular response to inflammatory stimuli leads to the production of eicosanoids—prostanoids (PRs) and leukotrienes (LTs)—and signaling molecules—cytokines and chemokines—by macrophages. Quantitative modeling of the inflammatory response is challenging owing to a lack of knowledge of the complex regulatory processes involved. Cybernetic models address these challenges by utilizing a well-defined cybernetic goal and optimizing a coarse-grained model toward this goal. We developed a cybernetic model to study arachidonic acid (AA) metabolism, which included two branches, PRs and LTs. We utilized a priori biological knowledge to define the branch-specific cybernetic goals for PR and LT branches as the maximization of TNFα and CCL2, respectively. We estimated the model parameters by fitting data from three experimental conditions. With these parameters, we were able to capture a novel fourth independent experimental condition as part of the model validation. The cybernetic model enhanced our understanding of enzyme dynamics by predicting their profiles. The success of the model implies that the cell regulates the synthesis and activity of the associated enzymes, through cybernetic control variables, to accomplish the chosen biological goal. The results indicated that the dominant metabolites are PGD2 (a PR) and LTB4 (an LT), aligning with their corresponding known prominent biological roles during inflammation. Using heuristic arguments, we also infer that eicosanoid overproduction can lead to increased secretion of cytokines/chemokines. This novel model integrates mechanistic knowledge, known biological understanding of signaling pathways, and data-driven methods to study the dynamics of eicosanoid metabolism.

Systemic approaches using single cell transcriptome reveal that C/EBPγ regulates autophagy under amino acid starved condition

Autophagy, a catabolic process to remove unnecessary or dysfunctional organelles, is triggered by various signals including nutrient starvation. Depending on the types of the nutrient deficiency, diverse sensing mechanisms and signaling pathways orchestrate for transcriptional and epigenetic regulation of autophagy. However, our knowledge about nutrient type-specific transcriptional regulation during autophagy is limited. To understand nutrient type-dependent transcriptional mechanisms during autophagy, we performed single cell RNA sequencing (scRNAseq) in the mouse embryonic fibroblasts (MEFs) with or without glucose starvation (GS) as well as amino acid starvation (AAS). Trajectory analysis using scRNAseq identified sequential induction of potential transcriptional regulators for each condition. Gene regulatory rules inferred using TENET newly identified CCAAT/enhancer binding protein γ (C/EBPγ) as a regulator of autophagy in AAS, but not GS, condition, and knockdown experiment confirmed the TENET result. Cell biological and biochemical studies validated that activating transcription factor 4 (ATF4) is responsible for conferring specificity to C/EBPγ for the activation of autophagy genes under AAS, but not under GS condition. Together, our data identified C/EBPγ as a previously unidentified key regulator under AAS-induced autophagy.

Multiplexed Visualization Method to Explore Complete Targeting Regulatory Relationships Among Circadian Genes for Insomnia Treatment

Understanding the complete map of melatonin synthesis, the information transfer network among circadian genes in pineal gland, promises to resolve outstanding issues in endocrine systems and improve the clinical diagnosis and treatment level of insomnia, immune disease and hysterical depression. Currently, some landmark studies have revealed some genes that regulate circadian rhythm associated with melatonin synthesis. However, these studies don't give a complete map of melatonin synthesis, as transfer information among circadian genes in pineal gland is lost. New biotechnology, integrates dynamic sequential omics and multiplexed imaging method, has been used to visualize the complete process of melatonin synthesis. It is found that there are two extremely significant information transfer processes involved in melatonin synthesis. In the first stage, as the light intensity decreased, melatonin synthesis mechanism has started, which is embodied in circadian genes, Rel, Polr2A, Mafk, and Srbf1 become active. In the second stage, circadian genes Hif1a, Bach1, Clock, E2f6, and Per2 are regulated simultaneously by four genes, Rel, Polr2A, Mafk, and Srbf1 and contribute genetic information to Aanat. The expeditious growth in this technique offer reference for an overall understanding of gene-to-gene regulatory relationship among circadian genes in pineal gland. In the study, dynamic sequential omics and the analysis process well provide the current state and future perspectives to better diagnose and cure diseases associated with melatonin synthesis disorder.

Inferring Gene Regulatory Networks From Single-Cell Transcriptomic Data Using Bidirectional RNN

Accurate inference of gene regulatory rules is critical to understanding cellular processes. Existing computational methods usually decompose the inference of gene regulatory networks (GRNs) into multiple subproblems, rather than detecting potential causal relationships simultaneously, which limits the application to data with a small number of genes. Here, we propose BiRGRN, a novel computational algorithm for inferring GRNs from time-series single-cell RNA-seq (scRNA-seq) data. BiRGRN utilizes a bidirectional recurrent neural network to infer GRNs. The recurrent neural network is a complex deep neural network that can capture complex, non-linear, and dynamic relationships among variables. It maps neurons to genes, and maps the connections between neural network layers to the regulatory relationship between genes, providing an intuitive solution to model GRNs with biological closeness and mathematical flexibility. Based on the deep network, we transform the inference of GRNs into a regression problem, using the gene expression data at previous time points to predict the gene expression data at the later time point. Furthermore, we adopt two strategies to improve the accuracy and stability of the algorithm. Specifically, we utilize a bidirectional structure to integrate the forward and reverse inference results and exploit an incomplete set of prior knowledge to filter out some candidate inferences of low confidence. BiRGRN is applied to four simulated datasets and three real scRNA-seq datasets to verify the proposed method. We perform comprehensive comparisons between our proposed method with other state-of-the-art techniques. These experimental results indicate that BiRGRN is capable of inferring GRN simultaneously from time-series scRNA-seq data. Our method BiRGRN is implemented in Python using the TensorFlow machine-learning library, and it is freely available at https://gitee.com/DHUDBLab/bi-rgrn.

Fusion prior gene network for high reliable single-cell gene regulatory network inference

Single-Cell RNA sequencing technology provides an opportunity to discover gene regulatory networks(GRN) that control cell differentiation and drive cell type transformation. However, it is faced with the challenge of high loss and high noise of sequencing data and contains many pseudo-connections. To solve these problems, we propose a framework called Fusion prior gene network for Gene Regulatory Network inference Accuracy Enhancement(FGRNAE) to infer a high reliable gene regulatory network. Specifically, based on the Single-Cell RNA-sequencing Network Propagation and network Fusion(scNPF) preprocessing framework, we employ the Random Walk with Restart on the prior gene network to interpolate the missing data. Furthermore, we infer the network using the Random Forest algorithm with the results achieved above. In addition, we apply data from the Co-Function Network to build a meta-gene network and select the regulatory connection with the Markov Random Field. Extensive experiments based on datasets from BEELINE validate the effectiveness of our framework for improving the accuracy of inference.

Assessing transfer entropy from biochemical data

We address the problem of evaluating the transfer entropy (TE) produced by biochemical reactions from experimentally measured data. Although these reactions are generally nonlinear and nonstationary processes making it challenging to achieve accurate modeling, Gaussian approximation can facilitate the TE assessment only by estimating covariance matrices using multiple data obtained from simultaneously measured time series representing the activation levels of biomolecules such as proteins. Nevertheless, the nonstationary nature of biochemical signals makes it difficult to theoretically assess the sampling distributions of TE, which are necessary for evaluating the statistical confidence and significance of the data-driven estimates. We resolve this difficulty by computationally assessing the sampling distributions using techniques from computational statistics. The computational methods are tested by using them in analyzing data generated from a theoretically tractable time-varying signal model, which leads to the development of a method to screen only statistically significant estimates. The usefulness of the developed method is examined by applying it to real biological data experimentally measured from the ERBB-RAS-MAPK system that superintends diverse cell fate decisions. A comparison between cells containing wild-type and mutant proteins exhibits a distinct difference in the time evolution of TE while any apparent difference is hardly found in average profiles of the raw signals. Such a comparison may help in unveiling important pathways of biochemical reactions.

Network inference with Granger causality ensembles on single-cell transcriptomics

Cellular gene expression changes throughout a dynamic biological process, such as differentiation. Pseudotimes estimate cells' progress along a dynamic process based on their individual gene expression states. Ordering the expression data by pseudotime provides information about the underlying regulator-gene interactions. Because the pseudotime distribution is not uniform, many standard mathematical methods are inapplicable for analyzing the ordered gene expression states. Here we present single-cell inference of networks using Granger ensembles (SINGE), an algorithm for gene regulatory network inference from ordered single-cell gene expression data. SINGE uses kernel-based Granger causality regression to smooth irregular pseudotimes and missing expression values. It aggregates predictions from an ensemble of regression analyses to compile a ranked list of candidate interactions between transcriptional regulators and target genes. In two mouse embryonic stem cell differentiation datasets, SINGE outperforms other contemporary algorithms. However, a more detailed examination reveals caveats about poor performance for individual regulators and uninformative pseudotimes.

VeTra: a tool for trajectory inference based on RNA velocity

MOTIVATION

Trajectory inference (TI) for single cell RNA sequencing (scRNAseq) data is a powerful approach to interpret dynamic cellular processes such as cell cycle and development. Still, however, accurate inference of trajectory is challenging. Recent development of RNA velocity provides an approach to visualize cell state transition without relying on prior knowledge.

RESULTS

To perform TI and group cells based on RNA velocity we developed VeTra. By applying cosine similarity and merging weakly connected components, VeTra identifies cell groups from the direction of cell transition. Besides, VeTra suggests key regulators from the inferred trajectory. VeTra is a useful tool for TI and subsequent analysis.

AVAILABILITY AND IMPLEMENTATION

The Vetra is available at https://github.com/wgzgithub/VeTra.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A Sight on Single-Cell Transcriptomics in Plants Through the Prism of Cell-Based Computational Modeling Approaches: Benefits and Challenges for Data Analysis

Single-cell technology is a relatively new and promising way to obtain high-resolution transcriptomic data mostly used for animals during the last decade. However, several scientific groups developed and applied the protocols for some plant tissues. Together with deeply-developed cell-resolution imaging techniques, this achievement opens up new horizons for studying the complex mechanisms of plant tissue architecture formation. While the opportunities for integrating data from transcriptomic to morphogenetic levels in a unified system still present several difficulties, plant tissues have some additional peculiarities. One of the plants' features is that cell-to-cell communication topology through plasmodesmata forms during tissue growth and morphogenesis and results in mutual regulation of expression between neighboring cells affecting internal processes and cell domain development. Undoubtedly, we must take this fact into account when analyzing single-cell transcriptomic data. Cell-based computational modeling approaches successfully used in plant morphogenesis studies promise to be an efficient way to summarize such novel multiscale data. The inverse problem's solutions for these models computed on the real tissue templates can shed light on the restoration of individual cells' spatial localization in the initial plant organ-one of the most ambiguous and challenging stages in single-cell transcriptomic data analysis. This review summarizes new opportunities for advanced plant morphogenesis models, which become possible thanks to single-cell transcriptome data. Besides, we show the prospects of microscopy and cell-resolution imaging techniques to solve several spatial problems in single-cell transcriptomic data analysis and enhance the hybrid modeling framework opportunities.

RVAgene: Generative modeling of gene expression time series data

Motivation

Methods to model dynamic changes in gene expression at a genome-wide level are not currently sufficient for large (temporally rich or single-cell) datasets. Variational autoencoders offer means to characterize large datasets and have been used effectively to characterize features of single-cell datasets. Here, we extend these methods for use with gene expression time series data.

Results

We present RVAgene: a recurrent variational autoencoder to model gene expression dynamics. RVAgene learns to accurately and efficiently reconstruct temporal gene profiles. It also learns a low dimensional representation of the data via a recurrent encoder network that can be used for biological feature discovery, and from which we can generate new gene expression data by sampling the latent space. We test RVAgene on simulated and real biological datasets, including embryonic stem cell differentiation and kidney injury response dynamics. In all cases, RVAgene accurately reconstructed complex gene expression temporal profiles. Via cross validation, we show that a low-error latent space representation can be learnt using only a fraction of the data. Through clustering and gene ontology term enrichment analysis on the latent space, we demonstrate the potential of RVAgene for unsupervised discovery. In particular, RVAgene identifies new programs of shared gene regulation of Lox family genes in response to kidney injury.

Availability and implementation

All datasets analyzed in this manuscript are publicly available and have been published previously. RVAgene is available in Python, at GitHub: https://github.com/maclean-lab/RVAgene; Zenodo archive: http://doi.org/10.5281/zenodo.4271097.

Supplementary information

Supplementary data are available at Bioinformatics online.