Inferring transcription factor regulatory networks from single-cell ATAC-seq data based on graph neural networks

Single-cell omics: experimental workflow, data analyses and applications

Cells are the fundamental units of biological systems and exhibit unique development trajectories and molecular features. Our exploration of how the genomes orchestrate the formation and maintenance of each cell, and control the cellular phenotypes of various organismsis, is both captivating and intricate. Since the inception of the first single-cell RNA technology, technologies related to single-cell sequencing have experienced rapid advancements in recent years. These technologies have expanded horizontally to include single-cell genome, epigenome, proteome, and metabolome, while vertically, they have progressed to integrate multiple omics data and incorporate additional information such as spatial scRNA-seq and CRISPR screening. Single-cell omics represent a groundbreaking advancement in the biomedical field, offering profound insights into the understanding of complex diseases, including cancers. Here, we comprehensively summarize recent advances in single-cell omics technologies, with a specific focus on the methodology section. This overview aims to guide researchers in selecting appropriate methods for single-cell sequencing and related data analysis.

Deep learning resilience inference for complex networked systems

Resilience, the ability to maintain fundamental functionality amidst failures and errors, is crucial for complex networked systems. Most analytical approaches rely on predefined equations for node activity dynamics and simplifying assumptions on network topology, limiting their applicability to real-world systems. Here, we propose ResInf, a deep learning framework integrating transformers and graph neural networks to infer resilience directly from observational data. ResInf learns representations of node activity dynamics and network topology without simplifying assumptions, enabling accurate resilience inference and low-dimensional visualization. Experimental results show that ResInf significantly outperforms analytical methods, with an F1-score improvement of up to 41.59% over Gao-Barzel-Barabási framework and 14.32% over spectral dimension reduction. It also generalizes to unseen topologies and dynamics and maintains robust performance despite observational disturbances. Our findings suggest that ResInf addresses an important gap in resilience inference for real-world systems, offering a fresh perspective on incorporating data-driven approaches to complex network modeling.

Artificial intelligence and machine learning applications for cultured meat

Cultured meat has the potential to provide a complementary meat industry with reduced environmental, ethical, and health impacts. However, major technological challenges remain which require time-and resource-intensive research and development efforts. Machine learning has the potential to accelerate cultured meat technology by streamlining experiments, predicting optimal results, and reducing experimentation time and resources. However, the use of machine learning in cultured meat is in its infancy. This review covers the work available to date on the use of machine learning in cultured meat and explores future possibilities. We address four major areas of cultured meat research and development: establishing cell lines, cell culture media design, microscopy and image analysis, and bioprocessing and food processing optimization. In addition, we have included a survey of datasets relevant to CM research. This review aims to provide the foundation necessary for both cultured meat and machine learning scientists to identify research opportunities at the intersection between cultured meat and machine learning.

Constructing the dynamic transcriptional regulatory networks to identify phenotype-specific transcription regulators

The transcriptional regulatory network (TRN) is a graph framework that helps understand the complex transcriptional regulation mechanisms in the transcription process. Identifying the phenotype-specific transcription regulators is vital to reveal the functional roles of transcription elements in associating the specific phenotypes. Although many methods have been developed towards detecting the phenotype-specific transcription elements based on the static TRN in the past decade, most of them are not satisfactory for elucidating the phenotype-related functional roles of transcription regulators in multiple levels, as the dynamic characteristics of transcription regulators are usually ignored in static models. In this study, we introduce a novel framework called DTGN to identify the phenotype-specific transcription factors (TFs) and pathways by constructing dynamic TRNs. We first design a graph autoencoder model to integrate the phenotype-oriented time-series gene expression data and static TRN to learn the temporal representations of genes. Then, based on the learned temporal representations of genes, we develop a statistical method to construct a series of dynamic TRNs associated with the development of specific phenotypes. Finally, we identify the phenotype-specific TFs and pathways from the constructed dynamic TRNs. Results from multiple phenotypic datasets show that the proposed DTGN framework outperforms most existing methods in identifying phenotype-specific TFs and pathways. Our framework offers a new approach to exploring the functional roles of transcription regulators that associate with specific phenotypes in a dynamic model.

Analyses of single-cell and bulk RNA sequencing combined with machine learning reveal the expression patterns of disrupted mitophagy in schizophrenia

Background

Mitochondrial dysfunction is an important factor in the pathogenesis of schizophrenia. However, the relationship between mitophagy and schizophrenia remains to be elucidated.

Methods

Single-cell RNA sequencing datasets of peripheral blood and brain organoids from SCZ patients and healthy controls were retrieved. Mitophagy-related genes that were differentially expressed between the two groups were screened. The diagnostic model based on key mitophagy genes was constructed using two machine learning methods, and the relationship between mitophagy and immune cells was analyzed. Single-cell RNA sequencing data of brain organoids was used to calculate the mitophagy score (Mitoscore).

Results

We found 7 key mitophagy genes to construct a diagnostic model. The mitophagy genes were related to the infiltration of neutrophils, activated dendritic cells, resting NK cells, regulatory T cells, resting memory T cells, and CD8 T cells. In addition, we identified 12 cell clusters based on the Mitoscore, and the most abundant neurons were further divided into three subgroups. Results at the single-cell level showed that Mitohigh_Neuron established a novel interaction with endothelial cells via SPP1 signaling pathway, suggesting their distinct roles in SCZ pathogenesis.

Conclusion

We identified a mitophagy signature for schizophrenia that provides new insights into disease pathogenesis and new possibilities for its diagnosis and treatment.

Inference and prioritization of tissue-specific regulons in Arabidopsis and Oryza

A regulon refers to a group of genes regulated by a transcription factor binding to regulatory motifs to achieve specific biological functions. To infer tissue-specific gene regulons in Arabidopsis, we developed a novel pipeline named InferReg. InferReg utilizes a gene expression matrix that includes 3400 Arabidopsis transcriptomes to make initial predictions about the regulatory relationships between transcription factors (TFs) and target genes (TGs) using co-expression patterns. It further improves these anticipated interactions by integrating TF binding site enrichment analysis to eliminate false positives that are only supported by expression data. InferReg further trained a graph convolutional network with 133 transcription factors, supported by ChIP-seq, as positive samples, to learn the regulatory logic between TFs and TGs to improve the accuracy of the regulatory network. To evaluate the functionality of InferReg, we utilized it to discover tissue-specific regulons in 5 Arabidopsis tissues: flower, leaf, root, seed, and seedling. We ranked the activities of regulons for each tissue based on reliability using Borda ranking and compared them with existing databases. The results demonstrated that InferReg not only identified known tissue-specific regulons but also discovered new ones. By applying InferReg to rice expression data, we were able to identify rice tissue-specific regulons, showing that our approach can be applied more broadly. We used InferReg to successfully identify important regulons in various tissues of Arabidopsis and Oryza, which has improved our understanding of tissue-specific regulations and the roles of regulons in tissue differentiation and development.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42994-024-00176-2.

Fast clustering and cell-type annotation of scATAC data using pre-trained embeddings

Data from the single-cell assay for transposase-accessible chromatin using sequencing (scATAC-seq) are now widely available. One major computational challenge is dealing with high dimensionality and inherent sparsity, which is typically addressed by producing lower dimensional representations of single cells for downstream clustering tasks. Current approaches produce such individual cell embeddings directly through a one-step learning process. Here, we propose an alternative approach by building embedding models pre-trained on reference data. We argue that this provides a more flexible analysis workflow that also has computational performance advantages through transfer learning. We implemented our approach in scEmbed, an unsupervised machine-learning framework that learns low-dimensional embeddings of genomic regulatory regions to represent and analyze scATAC-seq data. scEmbed performs well in terms of clustering ability and has the key advantage of learning patterns of region co-occurrence that can be transferred to other, unseen datasets. Moreover, models pre-trained on reference data can be exploited to build fast and accurate cell-type annotation systems without the need for other data modalities. scEmbed is implemented in Python and it is available to download from GitHub. We also make our pre-trained models available on huggingface for public use. scEmbed is open source and available at https://github.com/databio/geniml. Pre-trained models from this work can be obtained on huggingface: https://huggingface.co/databio.

CGMega: explainable graph neural network framework with attention mechanisms for cancer gene module dissection

Cancer is rarely the straightforward consequence of an abnormality in a single gene, but rather reflects a complex interplay of many genes, represented as gene modules. Here, we leverage the recent advances of model-agnostic interpretation approach and develop CGMega, an explainable and graph attention-based deep learning framework to perform cancer gene module dissection. CGMega outperforms current approaches in cancer gene prediction, and it provides a promising approach to integrate multi-omics information. We apply CGMega to breast cancer cell line and acute myeloid leukemia (AML) patients, and we uncover the high-order gene module formed by ErbB family and tumor factors NRG1, PPM1A and DLG2. We identify 396 candidate AML genes, and observe the enrichment of either known AML genes or candidate AML genes in a single gene module. We also identify patient-specific AML genes and associated gene modules. Together, these results indicate that CGMega can be used to dissect cancer gene modules, and provide high-order mechanistic insights into cancer development and heterogeneity.

Genome-wide nucleosome and transcription factor responses to genetic perturbations reveal chromatin-mediated mechanisms of transcriptional regulation

Epigenetic mechanisms contribute to gene regulation by altering chromatin accessibility through changes in transcription factor (TF) and nucleosome occupancy throughout the genome. Despite numerous studies focusing on changes in gene expression, the intricate chromatin-mediated regulatory code remains largely unexplored on a comprehensive scale. We address this by employing a factor-agnostic, reverse-genetics approach that uses MNase-seq to capture genome-wide TF and nucleosome occupancies in response to the individual deletion of 201 transcriptional regulators in Saccharomyces cerevisiae, thereby assaying nearly one million mutant-gene interactions. We develop a principled approach to identify and quantify chromatin changes genome-wide, observing differences in TF and nucleosome occupancy that recapitulate well-established pathways identified by gene expression data. We also discover distinct chromatin signatures associated with the up- and downregulation of genes, and use these signatures to reveal regulatory mechanisms previously unexplored in expression-based studies. Finally, we demonstrate that chromatin features are predictive of transcriptional activity and leverage these features to reconstruct chromatin-based transcriptional regulatory networks. Overall, these results illustrate the power of an approach combining genetic perturbation with high-resolution epigenomic profiling; the latter enables a close examination of the interplay between TFs and nucleosomes genome-wide, providing a deeper, more mechanistic understanding of the complex relationship between chromatin organization and transcription.

Inferring gene regulatory networks from single-cell multiome data using atlas-scale external data

Existing methods for gene regulatory network (GRN) inference rely on gene expression data alone or on lower resolution bulk data. Despite the recent integration of chromatin accessibility and RNA sequencing data, learning complex mechanisms from limited independent data points still presents a daunting challenge. Here we present LINGER (Lifelong neural network for gene regulation), a machine-learning method to infer GRNs from single-cell paired gene expression and chromatin accessibility data. LINGER incorporates atlas-scale external bulk data across diverse cellular contexts and prior knowledge of transcription factor motifs as a manifold regularization. LINGER achieves a fourfold to sevenfold relative increase in accuracy over existing methods and reveals a complex regulatory landscape of genome-wide association studies, enabling enhanced interpretation of disease-associated variants and genes. Following the GRN inference from reference single-cell multiome data, LINGER enables the estimation of transcription factor activity solely from bulk or single-cell gene expression data, leveraging the abundance of available gene expression data to identify driver regulators from case-control studies.

scButterfly: a versatile single-cell cross-modality translation method via dual-aligned variational autoencoders

Recent advancements for simultaneously profiling multi-omics modalities within individual cells have enabled the interrogation of cellular heterogeneity and molecular hierarchy. However, technical limitations lead to highly noisy multi-modal data and substantial costs. Although computational methods have been proposed to translate single-cell data across modalities, broad applications of the methods still remain impeded by formidable challenges. Here, we propose scButterfly, a versatile single-cell cross-modality translation method based on dual-aligned variational autoencoders and data augmentation schemes. With comprehensive experiments on multiple datasets, we provide compelling evidence of scButterfly's superiority over baseline methods in preserving cellular heterogeneity while translating datasets of various contexts and in revealing cell type-specific biological insights. Besides, we demonstrate the extensive applications of scButterfly for integrative multi-omics analysis of single-modality data, data enhancement of poor-quality single-cell multi-omics, and automatic cell type annotation of scATAC-seq data. Moreover, scButterfly can be generalized to unpaired data training, perturbation-response analysis, and consecutive translation.

EpiCarousel: memory- and time-efficient identification of metacells for atlas-level single-cell chromatin accessibility data

SUMMARY

Recent technical advancements in single-cell chromatin accessibility sequencing (scCAS) have brought new insights to the characterization of epigenetic heterogeneity. As single-cell genomics experiments scale up to hundreds of thousands of cells, the demand for computational resources for downstream analysis grows intractably large and exceeds the capabilities of most researchers. Here, we propose EpiCarousel, a tailored Python package based on lazy loading, parallel processing, and community detection for memory- and time-efficient identification of metacells, i.e. the emergence of homogenous cells, in large-scale scCAS data. Through comprehensive experiments on five datasets of various protocols, sample sizes, dimensions, number of cell types, and degrees of cell-type imbalance, EpiCarousel outperformed baseline methods in systematic evaluation of memory usage, computational time, and multiple downstream analyses including cell type identification. Moreover, EpiCarousel executes preprocessing and downstream cell clustering on the atlas-level dataset with 707 043 cells and 1 154 611 peaks within 2 h consuming <75 GB of RAM and provides superior performance for characterizing cell heterogeneity than state-of-the-art methods.

AVAILABILITY AND IMPLEMENTATION

The EpiCarousel software is well-documented and freely available at https://github.com/biox-nku/epicarousel. It can be seamlessly interoperated with extensive scCAS analysis toolkits.

DeepFGRN: inference of gene regulatory network with regulation type based on directed graph embedding

The inference of gene regulatory networks (GRNs) from gene expression profiles has been a key issue in systems biology, prompting many researchers to develop diverse computational methods. However, most of these methods do not reconstruct directed GRNs with regulatory types because of the lack of benchmark datasets or defects in the computational methods. Here, we collect benchmark datasets and propose a deep learning-based model, DeepFGRN, for reconstructing fine gene regulatory networks (FGRNs) with both regulation types and directions. In addition, the GRNs of real species are always large graphs with direction and high sparsity, which impede the advancement of GRN inference. Therefore, DeepFGRN builds a node bidirectional representation module to capture the directed graph embedding representation of the GRN. Specifically, the source and target generators are designed to learn the low-dimensional dense embedding of the source and target neighbors of a gene, respectively. An adversarial learning strategy is applied to iteratively learn the real neighbors of each gene. In addition, because the expression profiles of genes with regulatory associations are correlative, a correlation analysis module is designed. Specifically, this module not only fully extracts gene expression features, but also captures the correlation between regulators and target genes. Experimental results show that DeepFGRN has a competitive capability for both GRN and FGRN inference. Potential biomarkers and therapeutic drugs for breast cancer, liver cancer, lung cancer and coronavirus disease 2019 are identified based on the candidate FGRNs, providing a possible opportunity to advance our knowledge of disease treatments.

Prior knowledge-guided multilevel graph neural network for tumor risk prediction and interpretation via multi-omics data integration

The interrelation and complementary nature of multi-omics data can provide valuable insights into the intricate molecular mechanisms underlying diseases. However, challenges such as limited sample size, high data dimensionality and differences in omics modalities pose significant obstacles to fully harnessing the potential of these data. The prior knowledge such as gene regulatory network and pathway information harbors useful gene-gene interaction and gene functional module information. To effectively integrate multi-omics data and make full use of the prior knowledge, here, we propose a Multilevel-graph neural network (GNN): a hierarchically designed deep learning algorithm that sequentially leverages multi-omics data, gene regulatory networks and pathway information to extract features and enhance accuracy in predicting survival risk. Our method achieved better accuracy compared with existing methods. Furthermore, key factors nonlinearly associated with the tumor pathogenesis are prioritized by employing two interpretation algorithms (i.e. GNN-Explainer and IGscore) for neural networks, at gene and pathway level, respectively. The top genes and pathways exhibit strong associations with disease in survival analyses, many of which such as SEC61G and CYP27B1 are previously reported in the literature.

Gene regulatory network reconstruction: harnessing the power of single-cell multi-omic data

Inferring gene regulatory networks (GRNs) is a fundamental challenge in biology that aims to unravel the complex relationships between genes and their regulators. Deciphering these networks plays a critical role in understanding the underlying regulatory crosstalk that drives many cellular processes and diseases. Recent advances in sequencing technology have led to the development of state-of-the-art GRN inference methods that exploit matched single-cell multi-omic data. By employing diverse mathematical and statistical methodologies, these methods aim to reconstruct more comprehensive and precise gene regulatory networks. In this review, we give a brief overview on the statistical and methodological foundations commonly used in GRN inference methods. We then compare and contrast the latest state-of-the-art GRN inference methods for single-cell matched multi-omics data, and discuss their assumptions, limitations and opportunities. Finally, we discuss the challenges and future directions that hold promise for further advancements in this rapidly developing field.

Continuous lifelong learning for modeling of gene regulation from single cell multiome data by leveraging atlas-scale external data

Accurate context-specific Gene Regulatory Networks (GRNs) inference from genomics data is a crucial task in computational biology. However, existing methods face limitations, such as reliance on gene expression data alone, lower resolution from bulk data, and data scarcity for specific cellular systems. Despite recent technological advancements, including single-cell sequencing and the integration of ATAC-seq and RNA-seq data, learning such complex mechanisms from limited independent data points still presents a daunting challenge, impeding GRN inference accuracy. To overcome this challenge, we present LINGER (LIfelong neural Network for GEne Regulation), a novel deep learning-based method to infer GRNs from single-cell multiome data with paired gene expression and chromatin accessibility data from the same cell. LINGER incorporates both 1) atlas-scale external bulk data across diverse cellular contexts and 2) the knowledge of transcription factor (TF) motif matching to cis-regulatory elements as a manifold regularization to address the challenge of limited data and extensive parameter space in GRN inference. Our results demonstrate that LINGER achieves 2-3 fold higher accuracy over existing methods. LINGER reveals a complex regulatory landscape of genome-wide association studies, enabling enhanced interpretation of disease-associated variants and genes. Additionally, following the GRN inference from a reference sc-multiome data, LINGER allows for the estimation of TF activity solely from bulk or single-cell gene expression data, leveraging the abundance of available gene expression data to identify driver regulators from case-control studies. Overall, LINGER provides a comprehensive tool for robust gene regulation inference from genomics data, empowering deeper insights into cellular mechanisms.

STGRNS: an interpretable transformer-based method for inferring gene regulatory networks from single-cell transcriptomic data

MOTIVATION

Single-cell RNA-sequencing (scRNA-seq) technologies provide an opportunity to infer cell-specific gene regulatory networks (GRNs), which is an important challenge in systems biology. Although numerous methods have been developed for inferring GRNs from scRNA-seq data, it is still a challenge to deal with cellular heterogeneity.

RESULTS

To address this challenge, we developed an interpretable transformer-based method namely STGRNS for inferring GRNs from scRNA-seq data. In this algorithm, gene expression motif technique was proposed to convert gene pairs into contiguous sub-vectors, which can be used as input for the transformer encoder. By avoiding missing phase-specific regulations in a network, gene expression motif can improve the accuracy of GRN inference for different types of scRNA-seq data. To assess the performance of STGRNS, we implemented the comparative experiments with some popular methods on extensive benchmark datasets including 21 static and 27 time-series scRNA-seq dataset. All the results show that STGRNS is superior to other comparative methods. In addition, STGRNS was also proved to be more interpretable than "black box" deep learning methods, which are well-known for the difficulty to explain the predictions clearly.

AVAILABILITY AND IMPLEMENTATION

The source code and data are available at https://github.com/zhanglab-wbgcas/STGRNS.

Rise of Deep Learning Clinical Applications and Challenges in Omics Data: A Systematic Review

This research aims to review and evaluate the most relevant scientific studies about deep learning (DL) models in the omics field. It also aims to realize the potential of DL techniques in omics data analysis fully by demonstrating this potential and identifying the key challenges that must be addressed. Numerous elements are essential for comprehending numerous studies by surveying the existing literature. For example, the clinical applications and datasets from the literature are essential elements. The published literature highlights the difficulties encountered by other researchers. In addition to looking for other studies, such as guidelines, comparative studies, and review papers, a systematic approach is used to search all relevant publications on omics and DL using different keyword variants. From 2018 to 2022, the search procedure was conducted on four Internet search engines: IEEE Xplore, Web of Science, ScienceDirect, and PubMed. These indexes were chosen because they offer enough coverage and linkages to numerous papers in the biological field. A total of 65 articles were added to the final list. The inclusion and exclusion criteria were specified. Of the 65 publications, 42 are clinical applications of DL in omics data. Furthermore, 16 out of 65 articles comprised the review publications based on single- and multi-omics data from the proposed taxonomy. Finally, only a small number of articles (7/65) were included in papers focusing on comparative analysis and guidelines. The use of DL in studying omics data presented several obstacles related to DL itself, preprocessing procedures, datasets, model validation, and testbed applications. Numerous relevant investigations were performed to address these issues. Unlike other review papers, our study distinctly reflects different observations on omics with DL model areas. We believe that the result of this study can be a useful guideline for practitioners who look for a comprehensive view of the role of DL in omics data analysis.

Boosting single-cell gene regulatory network reconstruction via bulk-cell transcriptomic data

Computational recovery of gene regulatory network (GRN) has recently undergone a great shift from bulk-cell towards designing algorithms targeting single-cell data. In this work, we investigate whether the widely available bulk-cell data could be leveraged to assist the GRN predictions for single cells. We infer cell-type-specific GRNs from both the single-cell RNA sequencing data and the generic GRN derived from the bulk cells by constructing a weakly supervised learning framework based on the axial transformer. We verify our assumption that the bulk-cell transcriptomic data are a valuable resource, which could improve the prediction of single-cell GRN by conducting extensive experiments. Our GRN-transformer achieves the state-of-the-art prediction accuracy in comparison to existing supervised and unsupervised approaches. In addition, we show that our method can identify important transcription factors and potential regulations for Alzheimer's disease risk genes by using the predicted GRN. Availability: The implementation of GRN-transformer is available at https://github.com/HantaoShu/GRN-Transformer.

A benchmark study of deep learning-based multi-omics data fusion methods for cancer

BACKGROUND

A fused method using a combination of multi-omics data enables a comprehensive study of complex biological processes and highlights the interrelationship of relevant biomolecules and their functions. Driven by high-throughput sequencing technologies, several promising deep learning methods have been proposed for fusing multi-omics data generated from a large number of samples.

RESULTS

In this study, 16 representative deep learning methods are comprehensively evaluated on simulated, single-cell, and cancer multi-omics datasets. For each of the datasets, two tasks are designed: classification and clustering. The classification performance is evaluated by using three benchmarking metrics including accuracy, F1 macro, and F1 weighted. Meanwhile, the clustering performance is evaluated by using four benchmarking metrics including the Jaccard index (JI), C-index, silhouette score, and Davies Bouldin score. For the cancer multi-omics datasets, the methods' strength in capturing the association of multi-omics dimensionality reduction results with survival and clinical annotations is further evaluated. The benchmarking results indicate that moGAT achieves the best classification performance. Meanwhile, efmmdVAE, efVAE, and lfmmdVAE show the most promising performance across all complementary contexts in clustering tasks.

CONCLUSIONS

Our benchmarking results not only provide a reference for biomedical researchers to choose appropriate deep learning-based multi-omics data fusion methods, but also suggest the future directions for the development of more effective multi-omics data fusion methods. The deep learning frameworks are available at https://github.com/zhenglinyi/DL-mo .