Unraveling the relevance of SARS-Cov-2 infection and ferroptosis within the heart of COVID-19 patients.

Background: The coronavirus disease 2019 (COVID-19) was caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which led to a huge mortality rate and imposed significant costs on the health system, causing severe damage to the cells of different organs such as the heart. However, the exact details and mechanisms behind this damage are not clarified. Therefore, we aimed to identify the cell and molecular mechanism behind the heart damage caused by SARS-Cov-2 infection. Methods: RNA-seq data for COVID-19 patients' hearts was analyzed to obtain differentially expressed genes (DEGs) and differentially expressed ferroptosis-related genes (DEFRGs). Then, DEFRGs were used for analyzing GO and KEGG enrichment, and perdition of metabolites and drugs. we also constructed a PPI network and identified hub genes and functional modules for the DEFRGs. Subsequently, the hub genes were validated using two independent RNA-seq datasets. Finally, the miRNA-gene interaction networks were predicted in addition to a miRNA-TF co-regulatory network, and important miRNAs and transcription factors (TFs) were highlighted. Findings: We found ferroptosis transcriptomic alterations within the hearts of COVID-19 patients. The enrichment analyses suggested the involvement of DEFRGs in the citrate cycle pathway, ferroptosis, carbon metabolism, amino acid biosynthesis, and response to oxidative stress. IL6, CDH1, AR, EGR1, SIRT3, GPT2, VDR, PCK2, VDR, and MUC1 were identified as the ferroptosis-related hub genes. The important miRNAs and TFs were miR-124-3P, miR-26b-5p, miR-183-5p, miR-34a-5p and miR-155-5p; EGR1, AR, IL6, HNF4A, SRC, EZH2, PPARA, and VDR. Conclusion: These results provide a useful context and a cellular snapshot of how ferroptosis affects cardiomyocytes (CMs) in COVID-19 patients' hearts. Besides, suppressing ferroptosis seems to be a beneficial therapeutic approach to mitigate heart damage in COVID-19.

Regulatory Networks of Coresident Subgenomes during Rapid Fiber Cell Elongation in Upland Cotton.

Cotton, an intriguing plant species shaped by polyploidization, evolution, and domestication, holds particular interest due to the complex mechanisms governing fiber traits across its two subgenomes. However, the regulatory elements or transcriptional networks between subgenomes during fiber elongation remain elusive. Here, we analyzed 1,462 cotton fiber samples to reconstruct gene expression regulatory networks influencing fiber cell elongation. Inter-subgenomic eQTLs largely dictate gene transcription, with a notable tendency for the D subgenome to regulate A subgenome eGenes. This regulation showcases synchronized homoeologous gene expression driven by colocalized eQTLs and divergent patterns that diminish genetic correlations, thus leading to preferential expression in the A and D subgenomes. Hotspot456 emerged as a key regulator of fiber initiation and elongation, and artificial selection of trans-eQTLs in hotspot456 positively regulating KCS1 has facilitated cell elongation. To elucidate the roles of trans-eQTL in improved fiber breeding, experimentation confirmed the inhibition of GhTOL9 by a specific trans-eQTL via GhWRKY28, which negatively impacts fiber elongation. We propose a model where the GhWRKY28-GhTOL9 module, through the Endosomal Sorting Complex Required for Transport pathway, regulates this process. This research significantly advances our understanding of cotton's evolutionary, domestication processes, and the intricate regulatory mechanisms underlying significant plant traits.

Energy Landscape Reveals the Underlying Mechanism of Cancer-Adipose Conversion in Gene Network Models.

Cancer is a systemic heterogeneous disease involving complex molecular networks. Tumor formation involves an epithelial-mesenchymal transition (EMT), which promotes both metastasis and plasticity of cancer cells. Recent experiments have proposed that cancer cells can be transformed into adipocytes via a combination of drugs. However, the underlying mechanisms for how these drugs work, from a molecular network perspective, remain elusive. To reveal the mechanism of cancer-adipose conversion (CAC), this study adopts a systems biology approach by combing mathematical modeling and molecular experiments, based on underlying molecular regulatory networks. Four types of attractors are identified, corresponding to epithelial (E), mesenchymal (M), adipose (A) and partial/intermediate EMT (P) cell states on the CAC landscape. Landscape and transition path results illustrate that intermediate states play critical roles in the cancer to adipose transition. Through a landscape control approach, two new therapeutic strategies for drug combinations are identified, that promote CAC. These predictions are verified by molecular experiments in different cell lines. The combined computational and experimental approach provides a powerful tool to explore molecular mechanisms for cell fate transitions in cancer networks. The results reveal underlying mechanisms of intermediate cell states that govern the CAC, and identified new potential drug combinations to induce cancer adipogenesis.

The landscape of miRNA-mRNA regulatory network and cellular sources in inflammatory bowel diseases: insights from text mining and single cell RNA sequencing analysis.

Background: Inflammatory Bowel Diseases (IBDs), encompassing Ulcerative Colitis (UC) and Crohn's Disease (CD), are chronic, recurrent inflammatory conditions of the gastrointestinal tract. The microRNA (miRNA) -mRNA regulatory network is pivotal in the initiation and progression of IBDs. Although individual studies provide valuable insights into miRNA mechanisms in IBDs, they often have limited scope due to constraints in population diversity, sample size, sequencing platform variability, batch effects, and potential researcher bias. Our study aimed to construct comprehensive miRNA-mRNA regulatory networks and determine the cellular sources and functions of key miRNAs in IBD pathogenesis. Methods: To minimize potential bias from individual studies, we utilized a text mining-based approach on published scientific literature from PubMed and PMC databases to identify miRNAs and mRNAs associated with IBDs and their subtypes. We constructed miRNA-mRNA regulatory networks by integrating both predicted and experimentally validated results from DIANA, Targetscan, PicTar, Miranda, miRDB, and miRTarBase (all of which are databases for miRNA target annotation). The functions of miRNAs were determined through gene enrichment analysis of their target mRNAs. Additionally, we used two large-scale single-cell RNA sequencing datasets to identify the cellular sources of miRNAs and the association of their expression levels with clinical status, molecular and functional alternation in CD and UC. Results: Our analysis systematically summarized IBD-related genes using text-mining methodologies. We constructed three comprehensive miRNA-mRNA regulatory networks specific to IBD, CD, and UC. Through cross-analysis with two large-scale scRNA-seq datasets, we determined the cellular sources of the identified miRNAs. Despite originating from different cell types, hsa-miR-142, hsa-miR-145, and hsa-miR-146a were common to both CD and UC. Notably, hsa-miR-145 was identified as myofibroblast-specific in both CD and UC. Furthermore, we found that higher tissue repair and enhanced glucose and lipid metabolism were associated with hsa-miR-145 in myofibroblasts in both CD and UC contexts. Conclusion: This comprehensive approach revealed common and distinct miRNA-mRNA regulatory networks in CD and UC, identified cell-specific miRNA expressions (notably hsa-miR-145 in myofibroblasts), and linked miRNA expression to functional alterations in IBD. These findings not only enhance our understanding of IBD pathogenesis but also offer promising diagnostic biomarkers and therapeutic targets for clinical practice in managing IBDs.

Myotube Guidance: Shaping up the Musculoskeletal System.

Myofibers are highly specialized contractile cells of skeletal muscles, and dysregulation of myofiber morphogenesis is emerging as a contributing cause of myopathies and structural birth defects. Myotubes are the myofiber precursors and undergo a dramatic morphological transition into long bipolar myofibers that are attached to tendons on two ends. Similar to axon growth cones, myotube leading edges navigate toward target cells and form cell-cell connections. The process of myotube guidance connects myotubes with the correct tendons, orients myofiber morphology with the overall body plan, and generates a functional musculoskeletal system. Navigational signaling, addition of mass and volume, and identification of target cells are common events in myotube guidance and axon guidance, but surprisingly, the mechanisms regulating these events are not completely overlapping in myotubes and axons. This review summarizes the strategies that have evolved to direct myotube leading edges to predetermined tendon cells and highlights key differences between myotube guidance and axon guidance. The association of myotube guidance pathways with developmental disorders is also discussed.

Circulating miRNAs in the Plasma of Post-COVID-19 Patients with Typical Recovery and Those with Long-COVID Symptoms: Regulation of Immune Response-Associated Pathways.

Following the acute phase of SARS-CoV-2 infection, certain individuals experience persistent symptoms referred to as long COVID. This study analyzed the patients categorized into three distinct groups: (1) individuals presenting rheumatological symptoms associated with long COVID, (2) patients who have successfully recovered from COVID-19, and (3) donors who have never contracted COVID-19. A notable decline in the expression of miR-200c-3p, miR-766-3p, and miR-142-3p was identified among patients exhibiting rheumatological symptoms of long COVID. The highest concentration of miR-142-3p was found in healthy donors. One potential way to reduce miRNA concentrations is through antibody-mediated hydrolysis. Not only can antibodies possessing RNA-hydrolyzing activity recognize the miRNA substrate specifically, but they also catalyze its hydrolysis. The analysis of the catalytic activity of plasma antibodies revealed that antibodies from patients with long COVID demonstrated lower hydrolysis activity against five fluorescently labeled oligonucleotide sequences corresponding to the Flu-miR-146b-5p, Flu-miR-766-3p, Flu-miR-4742-3p, and Flu-miR-142-3p miRNAs and increased activity against the Flu-miR-378a-3p miRNA compared to other patient groups. The changes in miRNA concentrations and antibody-mediated hydrolysis of miRNAs are assumed to have a complex regulatory mechanism that is linked to gene pathways associated with the immune system. We demonstrate that all six miRNAs under analysis are associated with a large number of signaling pathways associated with immune response-associated pathways.

Leveraging gene correlations in single cell transcriptomic data.

BACKGROUND: Many approaches have been developed to overcome technical noise in single cell RNA-sequencing (scRNAseq). As researchers dig deeper into data-looking for rare cell types, subtleties of cell states, and details of gene regulatory networks-there is a growing need for algorithms with controllable accuracy and fewer ad hoc parameters and thresholds. Impeding this goal is the fact that an appropriate null distribution for scRNAseq cannot simply be extracted from data in which ground truth about biological variation is unknown (i.e., usually). RESULTS: We approach this problem analytically, assuming that scRNAseq data reflect only cell heterogeneity (what we seek to characterize), transcriptional noise (temporal fluctuations randomly distributed across cells), and sampling error (i.e., Poisson noise). We analyze scRNAseq data without normalization-a step that skews distributions, particularly for sparse data-and calculate p values associated with key statistics. We develop an improved method for selecting features for cell clustering and identifying gene-gene correlations, both positive and negative. Using simulated data, we show that this method, which we call BigSur (Basic Informatics and Gene Statistics from Unnormalized Reads), captures even weak yet significant correlation structures in scRNAseq data. Applying BigSur to data from a clonal human melanoma cell line, we identify thousands of correlations that, when clustered without supervision into gene communities, align with known cellular components and biological processes, and highlight potentially novel cell biological relationships. CONCLUSIONS: New insights into functionally relevant gene regulatory networks can be obtained using a statistically grounded approach to the identification of gene-gene correlations.

Tissue-specific RNA-seq defines genes governing male tail tip morphogenesis in C. elegans.

Caenorhabditis elegans males undergo sex-specific tail tip morphogenesis (TTM) under the control of the DM-domain transcription factor DMD-3. To find genes regulated by DMD-3, we performed RNA-seq of laser-dissected tail tips. We identified 564 genes differentially expressed (DE) in wild-type males versus dmd-3(-) males and hermaphrodites. The transcription profile of dmd-3(-) tail tips is similar to that in hermaphrodites. For validation, we analyzed transcriptional reporters for 49 genes and found male-specific or male-biased expression for 26 genes. Only 11 DE genes overlapped with genes found in a previous RNAi screen for defective TTM. GO enrichment analysis of DE genes finds upregulation of genes within the unfolded protein response pathway and downregulation of genes involved in cuticle maintenance. Of the DE genes, 40 are transcription factors, indicating that the gene network downstream of DMD-3 is complex and potentially modular. We propose modules of genes that act together in TTM and are co-regulated by DMD-3, among them the chondroitin synthesis pathway and the hypertonic stress response.

Dissecting reversible and irreversible single cell state transitions from gene regulatory networks.

Understanding cell state transitions and their governing regulatory mechanisms remains one of the fundamental questions in biology. We develop a computational method, state transition inference using cross-cell correlations (STICCC), for predicting reversible and irreversible cell state transitions at single-cell resolution by using gene expression data and a set of gene regulatory interactions. The method is inspired by the fact that the gene expression time delays between regulators and targets can be exploited to infer past and future gene expression states. From applications to both simulated and experimental single-cell gene expression data, we show that STICCC-inferred vector fields capture basins of attraction and irreversible fluxes. By connecting regulatory information with systems' dynamical behaviors, STICCC reveals how network interactions influence reversible and irreversible state transitions. Compared to existing methods that infer pseudotime and RNA velocity, STICCC provides complementary insights into the gene regulation of cell state transitions.

Inferring gene regulatory networks with graph convolutional network based on causal feature reconstruction.

Inferring gene regulatory networks through deep learning and causal inference methods is a crucial task in the field of computational biology and bioinformatics. This study presents a novel approach that uses a Graph Convolutional Network (GCN) guided by causal information to infer Gene Regulatory Networks (GRN). The transfer entropy and reconstruction layer are utilized to achieve causal feature reconstruction, mitigating the information loss problem caused by multiple rounds of neighbor aggregation in GCN, resulting in a causal and integrated representation of node features. Separable features are extracted from gene expression data by the Gaussian-kernel Autoencoder to improve computational efficiency. Experimental results on the DREAM5 and the mDC dataset demonstrate that our method exhibits superior performance compared to existing algorithms, as indicated by the higher values of the AUPRC metrics. Furthermore, the incorporation of causal feature reconstruction enhances the inferred GRN, rendering them more reasonable, accurate, and reliable.

Definition of regulatory elements and transcription factors controlling porcine immune cell gene expression at single cell resolution using single nucleus ATAC-seq.

The transcriptome of porcine peripheral blood mononuclear cells (PBMC) at single cell (sc) resolution is well described, but little is understood about the cis-regulatory mechanism behind scPBMC gene expression. Here, we profiled the open chromatin landscape of porcine PBMC that define cis-regulatory elements and mechanism contributing to the transcription using single nucleus ATAC sequencing (snATAC-seq). Approximately 22 % of the identified peaks overlapped with annotated transcription start sites (TSS). Using clustering based on open chromatin pattern similarity, we demonstrate that cell type annotations using snATAC-seq are highly concordant to that reported for sc RNA sequencing (scRNA-seq). The differentially accessible peaks (DAPs) for each cell type were characterized and the pattern of accessibility of the DAPs near cell type markers across cell types was similar to that of the average gene expression level of corresponding marker genes. Additionally, we found that peaks identified in snATAC-seq have the potential power to predict the cell type specific transcription starting site (TSS). We identified both transcription factors (TFs) whose binding motif were enriched in cell type DAPs of multiple cell types and cell type specific TFs by conducting transcription factor binding motif (TFBM) analysis. Furthermore, we identified the putative enhancer or promoter regions bound by TFs for each differentially expressed gene (DEG) with a DAP that overlapped with its TSS by generating cis-co-accessibility networks (CCAN). To predict the regulators of such DEGs, TFBM analysis was performed for each CCAN. The regulator TF-target DEG pairs predicted in this way were largely consistent with the results reported in the ENCODE Transcription Factor Targets Dataset (TFTD). This snATAC-seq approach provides insights into the regulation of chromatin accessibility landscape of porcine PBMCs and enables discovery of TFs predicted to control DEG through binding regulatory elements whose chromatin accessibility correlates with the DEG promoter region.

Non-coding RNA and its network in the pathogenesis of Myasthenia Gravis.

Myasthenia Gravis (MG) is a chronic autoimmune disease that primarily affects the neuromuscular junction, leading to muscle weakness in patients with this condition. Previous studies have identified several dysfunctions in thymus and peripheral blood mononuclear cells (PBMCs), such as the formation of ectopic germinal centers in the thymus and an imbalance of peripheral T helper cells and regulatory T cells, that contribute to the initiation and development of MG. Recent evidences suggest that noncoding RNA, including miRNA, lncRNA and circRNA may play a significant role in MG progression. Additionally, the network between these noncoding RNAs, such as the competing endogenous RNA regulatory network, has been found to be involved in MG progression. In this review, we summarized the roles of miRNA, lncRNA, and circRNA, highlighted their potential application as biomarkers in diagnosing MG, and discussed their potential regulatory networks in the abnormal thymus and PBMCs during MG development.

Identification of VvAGL Genes Reveals Their Network's Involvement in the Modulation of Seed Abortion via Responding Multi-Hormone Signals in Grapevines.

The formation of seedless traits is regulated by multiple factors. AGLs, which belong to the MADS-box family, were reported to be important regulators in this process; however, the underlying mechanism remains elusive. Here, we identified the VvAGL sub-family genes during the seed abortion process in seedless grapevine cv. 'JingkeJing' and found 40 differentially expressed VvAGL members and 1069 interacting proteins in this process. Interestingly, almost all members and their interacting proteins involved in the tryptophan metabolic pathway (K14486) and participated in the phytohormone signalling (KO04075) pathway, including the growth hormone (IAA), salicylic acid (SA), abscisic acid (ABA), cytokinin (CTK), and ethylene signalling pathways. The promoters of AGL sub-family genes contain cis-elements in response to hormones such as IAA, ABA, CTK, SA, and ETH, implying that they might respond to multi-hormone signals and involve in hormone signal transductions. Further expression analysis revealed VvAGL6-2, VvAGL11, VvAGL62-11, and VvAGL15 had the highest expression at the critical period of seed abortion, and there were positive correlations between ETH-VvAGL15-VvAGL6-2, ABA-VvAGL80, and SA-VvAGL62 in promoting seed abortion but negative feedback between IAA-VvAGL15-VvAGL6-2 and CTK-VvAGL11. Furthermore, many genes in the IAA, ABA, SA, CTK, and ETH pathways had a special expressional pattern in the seed, whereby we developed a regulatory network mediated by VvAGLs by responding to multihormonal crosstalk during grape seed abortion. Our findings provide new insights into the regulatory network of VvAGLs in multi-hormone signalling to regulate grape seed abortion, which could be helpful in the molecular breeding of high-quality seedless grapes.

Employ machine learning to identify NAD+ metabolism-related diagnostic markers for ischemic stroke and develop a diagnostic model.

Ischemic stroke (IS) is a severe condition regulated by complex molecular alterations. This study aimed to identify potential nicotinamide adenine dinucleotide (NAD+) metabolism-associated diagnostic markers of IS and explore their associations with immune dynamics. Weighted Gene Co-expression Network Analysis and single-sample gene set enrichment analysis (ssGSEA) were employed to identify key gene modules on the GEO dataset (GSE16561). LASSO regression was used to identify diagnostic genes. A diagnostic model was then developed using the training dataset, and its performance was assessed using a validation dataset (GSE22255 dataset). Associations between hub genes and immune cells, immune response genes, and human leukocyte antigen (HLA) genes were assessed by ssGSEA. A regulatory network was constructed using mirBase and TRRUST databases. A total of 20 NAD+ metabolic genes exhibited noteworthy expression variations. Within the module notably associated with NAD+ metabolism, 19 specific genes were included in the diagnostic model, which was validated on the GSE22255 dataset (AUC: 0.733). There were significant disparities in immune cell populations, immune response genes, and HLA gene expression, all of which were associated with the hub genes. A regulatory network composed of 153 edges and 103 nodes was constructed. This study advances our understanding of IS by providing insights into NAD+ metabolism and gene interactions, contributing to potential diagnostic innovations in IS.

GRNMOPT: Inference of gene regulatory networks based on a multi-objective optimization approach.

BACKGROUND AND OBJECTIVE: The reconstruction of gene regulatory networks (GRNs) stands as a vital approach in deciphering complex biological processes. The application of nonlinear ordinary differential equations (ODEs) models has demonstrated considerable efficacy in predicting GRNs. Notably, the decay rate and time delay are pivotal in authentic gene regulation, yet their systematic determination in ODEs models remains underexplored. The development of a comprehensive optimization framework for the effective estimation of these key parameters is essential for accurate GRN inference. METHOD: This study introduces GRNMOPT, an innovative methodology for inferring GRNs from time-series and steady-state data. GRNMOPT employs a combined use of decay rate and time delay in constructing ODEs models to authentically represent gene regulatory processes. It incorporates a multi-objective optimization approach, optimizing decay rate and time delay concurrently to derive Pareto optimal sets for these factors, thereby maximizing accuracy metrics such as AUROC (Area Under the Receiver Operating Characteristic curve) and AUPR (Area Under the Precision-Recall curve). Additionally, the use of XGBoost for calculating feature importance aids in identifying potential regulatory gene links. RESULTS: Comprehensive experimental evaluations on two simulated datasets from DREAM4 and three real gene expression datasets (Yeast, In vivo Reverse-engineering and Modeling Assessment [IRMA], and Escherichia coli [E. coli]) reveal that GRNMOPT performs commendably across varying network scales. Furthermore, cross-validation experiments substantiate the robustness of GRNMOPT. CONCLUSION: We propose a novel approach called GRNMOPT to infer GRNs based on a multi-objective optimization framework, which effectively improves inference accuracy and provides a powerful tool for GRNs inference.

Prognostic value and molecular mechanisms of OAS1 in lung adenocarcinoma.

BACKGROUND: The expression of 2'-5'-oligoadenylate synthetase 1 (OAS1) in lung cancer has been validated in numerous studies. However, the prognostic value of OAS1 expression in lung adenocarcinoma (LUAD) still remains unclear. This study aimed to reveal the prognostic value and associated molecular mechanisms of OAS1 expression in LUAD. METHODS: Gene expression data of LUAD were extracted from online databases. Gene and protein expression levels of OAS1 in LUAD and normal samples were revealed, followed by prognostic analysis of OAS1. Next, we conducted a thorough bioinformatics analysis to examine the enrichment of key functional and biological signaling pathways and their correlation with the abundance of immune cells. The independent prognoses, drug responses, and PPI networks associated with OAS1 were analyzed. OAS1 expression was evaluated in LUAD tissues and cell lines. OAS1 was knocked down by siRNA transfection, followed by CCK8, colony formation, and wound-healing assays. RESULTS: Gene and protein expression levels of OAS1 in LUAD samples were significantly higher than those in normal samples (all P < 0.05). OAS1 stimulation were correlated with poor prognosis, lymph node metastasis, advanced tumor stage, immune cells, and immunomodulators. The prognostic value of OAS1 in LUAD was determined via univariate regression analysis. In total, 10 OAS1-associated genes were revealed via PPI analysis of OAS1, which were primarily enriched in functions, such as the negative regulation of viral genome replication. Transcriptional analysis revealed several OAS1-related interactions, including STAT3-miR-21-OAS1. STAT3 was overexpressed and miR-21 was expressed in LUAD cells. Upregulation of OAS1 protein was determined in LUAD tissues and cell lines. OAS1 knockdown significantly reduced proliferation and migration of LUAD cells. CONCLUSIONS: OAS1 overexpression influenced survival and immune cell infiltration in patients with LUAD, which might be a potential prognostic gene for LUAD. Moreover, OAS1 contributed to LUAD progression by participating in STAT3-miR-21-OAS1 axis.

BpfD Is a c-di-GMP Effector Protein Playing a Key Role for Pellicle Biosynthesis in Shewanella oneidensis.

The aquatic γ-proteobacterium Shewanella oneidensis is able to form two types of biofilms: a floating biofilm at the air-liquid interface (pellicle) and a solid surface-associated biofilm (SSA-biofilm). S. oneidensis possesses the Bpf system, which is orthologous to the Lap system first described in Pseudomonas fluorescens. In the Lap systems, the retention of a large adhesin (LapA) at the cell surface is controlled by LapD, a c-di-GMP effector protein, and LapG, a periplasmic protease targeting LapA. Here, we showed that the Bpf system is mandatory for pellicle biogenesis, but not for SSA-biofilm formation, indicating that the role of Bpf is somewhat different from that of Lap. The BpfD protein was then proved to bind c-di-GMP via its degenerated EAL domain, thus acting as a c-di-GMP effector protein like its counterpart LapD. In accordance with its key role in pellicle formation, BpfD was found to interact with two diguanylate cyclases, PdgA and PdgB, previously identified as involved in pellicle formation. Finally, BpfD was shown to interact with CheY3, the response regulator controlling both chemotaxis and biofilm formation. Altogether, these results indicate that biofilm formation in S. oneidensis is under the control of a large c-di-GMP network.

Gene regulatory network inference based on novel ensemble method.

Gene regulatory networks (GRNs) contribute toward understanding the function of genes and the development of cancer or the impact of key genes on diseases. Hence, this study proposes an ensemble method based on 13 basic classification methods and a flexible neural tree (FNT) to improve GRN identification accuracy. The primary classification methods contain ridge classification, stochastic gradient descent, Gaussian process classification, Bernoulli Naive Bayes, adaptive boosting, gradient boosting decision tree, hist gradient boosting classification, eXtreme gradient boosting (XGBoost), multilayer perceptron, light gradient boosting machine, random forest, support vector machine, and k-nearest neighbor algorithm, which are regarded as the input variable set of FNT model. Additionally, a hybrid evolutionary algorithm based on a gene programming variant and particle swarm optimization is developed to search for the optimal FNT model. Experiments on three simulation datasets and three real single-cell RNA-seq datasets demonstrate that the proposed ensemble feature outperforms 13 supervised algorithms, seven unsupervised algorithms (ARACNE, CLR, GENIE3, MRNET, PCACMI, GENECI, and EPCACMI) and four single cell-specific methods (SCODE, BiRGRN, LEAP, and BiGBoost) based on the area under the receiver operating characteristic curve, area under the precision-recall curve, and F1 metrics.

Machine learning reveals the transcriptional regulatory network and circadian dynamics of Synechococcus elongatus PCC 7942.

Synechococcus elongatus is an important cyanobacterium that serves as a versatile and robust model for studying circadian biology and photosynthetic metabolism. Its transcriptional regulatory network (TRN) is of fundamental interest, as it orchestrates the cell's adaptation to the environment, including its response to sunlight. Despite the previous characterization of constituent parts of the S. elongatus TRN, a comprehensive layout of its topology remains to be established. Here, we decomposed a compendium of 300 high-quality RNA sequencing datasets of the model strain PCC 7942 using independent component analysis. We obtained 57 independently modulated gene sets, or iModulons, that explain 67% of the variance in the transcriptional response and 1) accurately reflect the activity of known transcriptional regulations, 2) capture functional components of photosynthesis, 3) provide hypotheses for regulon structures and functional annotations of poorly characterized genes, and 4) describe the transcriptional shifts under dynamic light conditions. This transcriptome-wide analysis of S. elongatus provides a quantitative reconstruction of the TRN and presents a knowledge base that can guide future investigations. Our systems-level analysis also provides a global TRN structure for S. elongatus PCC 7942.

Machine-Learning Analysis of Streptomyces coelicolor Transcriptomes Reveals a Transcription Regulatory Network Encompassing Biosynthetic Gene Clusters.

Streptomyces produces diverse secondary metabolites of biopharmaceutical importance, yet the rate of biosynthesis of these metabolites is often hampered by complex transcriptional regulation. Therefore, a fundamental understanding of transcriptional regulation in Streptomyces is key to fully harness its genetic potential. Here, independent component analysis (ICA) of 454 high-quality gene expression profiles of the model species Streptomyces coelicolor is performed, of which 249 profiles are newly generated for S. coelicolor cultivated on 20 different carbon sources and 64 engineered strains with overexpressed sigma factors. ICA of the transcriptome dataset reveals 117 independently modulated groups of genes (iModulons), which account for 81.6% of the variance in the dataset. The genes in each iModulon are involved in specific cellular responses, which are often transcriptionally controlled by specific regulators. Also, iModulons accurately predict 25 secondary metabolite biosynthetic gene clusters encoded in the genome. This systemic analysis leads to reveal the functions of previously uncharacterized genes, putative regulons for 40 transcriptional regulators, including 30 sigma factors, and regulation of secondary metabolism via phosphate- and iron-dependent mechanisms in S. coelicolor. ICA of large transcriptomic datasets thus enlightens a new and fundamental understanding of transcriptional regulation of secondary metabolite synthesis along with interconnected metabolic processes in Streptomyces.