Integrating Gene Expression Profiles Research Articles

Society of Surgical Oncology Consensus Statement: Assessing the Evidence for and Utility of Gene Expression Profiling of Primary Cutaneous Melanoma.

Gene expression profiling (GEP) of primary cutaneous melanoma aims to offer prognostic and predictive information to guide clinical care. Despite limited evidence of clinical utility, these tests are increasingly incorporated into clinical care. A panel of melanoma experts from the Society of Surgical Oncology convened to develop recommendations regarding the use of GEP to guide management of patients with melanoma. The use of currently available GEP tests were evaluated in three clinical scenarios: (1) the utility in patient selection for sentinel lymph node biopsy; (2) the utility to guide surveillance; and (3) the utility to inform adjuvant therapy. As a basis for these recommendations, the panel performed a systematic review of the literature, including articles published from January 2012 until August 2023. After review of 137 articles,50met the inclusion criteria. These articles included evidence related to three available GEP tests: 31-GEP, CP-GEP, and 11-GEP. The consensus recommendations were finalized using a modified Delphi process. The panel found that current evidence often fails to account for known clinicopathologic risk factors and lacks high-level data. The panel recognizes that thestudy of GEP tests is still evolving. The integration of GEP into routine clinical practice for predicting sentinel lymph node status and patient prognosis in melanoma is therefore not currently recommended. At present, GEP should be considered primarily an investigational tool, ideally used in the context of clinical trials or specialized research settings.

STASCAN deciphers fine-resolution cell distribution maps in spatial transcriptomics by deep learning

Spatial transcriptomics technologies have been widely applied to decode cellular distribution by resolving gene expression profiles in tissue. However, sequencing techniques still limit the ability to create a fine-resolved spatial cell-type map. To this end, we develop a novel deep-learning-based approach, STASCAN, to predict the spatial cellular distribution of captured or uncharted areas where only histology images are available by cell feature learning integrating gene expression profiles and histology images. STASCAN is successfully applied across diverse datasets from different spatial transcriptomics technologies and displays significant advantages in deciphering higher-resolution cellular distribution and resolving enhanced organizational structures.

Exploring T-cell exhaustion features in Acute myocardial infarction for a Novel Diagnostic model and new therapeutic targets by bio-informatics and machine learning

BackgroundT-cell exhaustion (TEX), a condition characterized by impaired T-cell function, has been implicated in numerous pathological conditions, but its role in acute myocardial Infarction (AMI) remains largely unexplored. This research aims to identify and characterize all TEX-related genes for AMI diagnosis.MethodsBy integrating gene expression profiles, differential expression analysis, gene set enrichment analysis, protein-protein interaction networks, and machine learning algorithms, we were able to decipher the molecular mechanisms underlying TEX and its significant association with AMI. In addition, we investigated the diagnostic validity of the leading TEX-related genes and their interactions with immune cell profiles. Different types of candidate small molecule compounds were ultimately matched with TEX-featured genes in the “DrugBank” database to serve as potential therapeutic medications for future TEX-AMI basic research.ResultsWe screened 1725 differentially expressed genes (DEGs) from 80 AMI samples and 71 control samples, identifying 39 differential TEX-related transcripts in total. Functional enrichment analysis identified potential biological functions and signaling pathways associated with the aforementioned genes. We constructed a TEX signature containing five hub genes with favorable prognostic performance using machine learning algorithms. In addition, the prognostic performance of the nomogram of these five hub genes was adequate (AUC between 0.815 and 0.995). Several dysregulated immune cells were also observed. Finally, six small molecule compounds which could be the future therapeutic for TEX in AMI were discovered.ConclusionFive TEX diagnostic feature genes, CD48, CD247, FCER1G, TNFAIP3, and FCGRA, were screened in AMI. Combining these genes may aid in the early diagnosis and risk prediction of AMI, as well as the evaluation of immune cell infiltration and the discovery of new therapeutics.

Abstract 6395: Metabolic dysregulation as a determinant of prognosis and response to sorafenib in liver cancer

Abstract Background & Aims: Elevated metabolic activity is a defining characteristic of liver cancer. Nevertheless, our understanding of the molecular underpinnings of high metabolic activity in liver cancer and its impact on clinical outcomes and treatment responses remains limited. We aimed to elucidate the metabolic features of liver cancer associated with clinical outcomes, particularly overall survival and responsiveness to sorafenib treatment. Methods: Utilizing cross-species comparisons of genomic data, we integrated gene expression profiles from liver tissues in metabolically challenged mice and from patients with cirrhosis and liver cancer. To categorize patients, we developed a genomic predictor and applied it to gene expression data from a cirrhosis cohort (n=216) and large liver cancer cohorts (n=1390). Comprehensive statistical and informatics analyses were conducted to evaluate clinical significance, including overall survival and responses to standard treatments. Moreover, we integrated genomic data derived from patient-derived xenograft (PDX) models into our analytical framework. Results: Our integrated data analysis unveiled three distinct metabolic subtypes of tissues: high, moderate, and low metabolic subtypes. In the cirrhosis cohort, the high metabolic subtype was significantly associated with early liver cancer development. Among liver cancer cohorts, the high metabolic subtype exhibited characteristics such as poor survival, heightened hepatic stem cell features, increased genomic instability, and elevated expression of AFP, KRT19, EPCAM, and SALL4. Notably, when liver cancer patients undergoing systemic sorafenib treatment were stratified using our genomic predictor, those with low and moderate metabolic activity derived significant benefits from sorafenib therapy. Conversely, the high metabolic subtype displayed substantial resistance to sorafenib, suggesting that metabolic activity may dictate the response of liver cancer cells to sorafenib treatment. Gene network analysis of the metabolic expression signature identified multiple signaling pathways that potentially contribute to sorafenib resistance in liver cancer cells. Our analysis predicted the activation of pathways such as the TGF-β pathway (TGFB1), consistent with previous findings of TGF-β pathway activation in highly metabolic cancer cells. Additionally, our analysis identified YAP1, HIF1A, and MAPK pathways as activated pathways in sorafenib-resistant liver cancer. Conclusions: This study identifies clinically and metabolically distinct subtypes of liver cancer, biomarkers associated with these subtypes, and mechanisms underlying metabolic-mediated resistance of liver cancer cells to sorafenib. Furthermore, the metabolic subtypes observed in PDX models provide valuable tools for selecting appropriate preclinical models for future research endeavors. Citation Format: Hyewon Park, Sowon Park, Sung-Hwan Lee, Yun Seong Jeong, Bohwa Sohn, Sun Young Yim, Ju-Seog Lee. Metabolic dysregulation as a determinant of prognosis and response to sorafenib in liver cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6395.

The cell death-related genes machine learning model for precise therapy and clinical drug selection in hepatocellular carcinoma.

Hepatocellular carcinoma (HCC) is the prevailing subtype of hepatocellular malignancy. While previous investigations have evidenced a robust link with programmed cell death (PCD) and tumorigenesis, a comprehensive inquiry targeting the relationship between multiple PCDs and HCC remains scant. Our aim was to develop a predictive model for different PCD patterns in order to investigate their impact on survival rates, prognosis and drug response rates in HCC patients. We performed functional annotation and pathway analysis on identified PCD-related genes (PCDRGs) using multiple bioinformatics tools. The prognostic value of these PCDRGs was verified through a dataset obtained from GEO. Consensus clustering analysis was utilized to elucidate the correlation between diverse PCD clusters and pertinent clinical characteristics. To comprehensively uncover the distinct PCD regulatory patterns, our analysis integrated gene expression profiling, immune cell infiltration and enrichment analysis. To predict survival differences in HCC patients, we established a PCD model. To enhance the clinical applicability for the model, we developed a highly accurate nomogram. To address the treatment of HCC, we identified several promising chemotherapeutic agents and novel targeted drugs. These drugs may be effective in treating HCC and could improve patient outcomes. To develop a cell death feature for HCC patients, we conducted an analysis of 12 different PCD mechanisms using eligible data obtained from public databases. Through this analysis, we were able to identify 1254 PCDRGs likely to contribute to cell death on HCC. Further analysis of 1254 PCDRGs identified 37 genes with prognostic value in HCC patients. These genes were then categorized into two PCD clusters A and B. The categorization was based on the expression patterns of the genes in the different clusters. Patients in PCD cluster B had better survival probabilities. This suggests that PCD mechanisms, as represented by the genes in cluster B, may have a protective effect against HCC progression. Furthermore, the expression of PCDRGs was significantly higher in PCD cluster A, indicating that this cluster may be more closely associated with PCD mechanisms. Furthermore, our observations indicate that patients exhibiting elevated tumour mutation burden (TMB) are at an augmented risk of mortality, in comparison to those displaying low TMB and low-risk statuses, who are more likely to experience prolonged survival. In addition, we have investigated the potential distinctions in the susceptibility of diverse risk cohorts towards emerging targeted therapies, designed for the treatment of HCC. Moreover, our investigation has shown that AZD2014, SB505124, LJI308 and OSI-207 show a greater efficacy in patients in the low-risk category. Conversely, for the high-risk group patients, PD173074, ZM447439 and CZC24832 exhibit a stronger response. Our findings suggest that the identification of risk groups and personalized treatment selection could lead to better clinical outcomes for patients with HCC. Furthermore, significant heterogeneity in clinical response to ICI therapy was observed among HCC patients with varying PCD expression patterns. This novel discovery underscores the prospective usefulness of these expression patterns as prognostic indicators for HCC patients and may aid in tailoring targeted treatment for those of distinct risk strata. Our investigation introduces a novel prognostic model for HCC that integrates diverse PCD expression patterns. This innovative model provides a novel approach for forecasting prognosis and assessing drug sensitivity in HCC patients, driving a more personalized and efficacious treatment paradigm, elevating clinical outcomes. Nonetheless, additional research endeavours are required to confirm the model's precision and assess its potential to inform clinical decision-making for HCC patients.

Blood transcriptome analysis uncovered COVID-19–myocarditis crosstalk

BackgroundThe condition of COVID-19-related myocarditis has emerged as a prominent contributor to COVID-19 mortality. As the epidemic persists, its incidence continues to rise. Despite ongoing efforts, the elucidation of COVID-19-related myocarditis underlying molecular mechanisms still requires further investigation. MethodsHub genes for COVID-19-related myocarditis were screened by integrating gene expression profile analysis via differential expression in COVID-19 (GSE196822) and myocarditis (GSE148153 and GSE147517). After verification with independent datasets (GSE211979, GSE167028, GSE178491 and GSE215865), the hub genes were studied using a range of systems-biology approaches, such as ceRNA, TF-mRNA networks and PPI networks, as well as gene ontology, pathway enrichment, immune infiltration analysis and drug target identification. ResultsTBKBP1 and ERGIC1 were identified as COVID-19-related myocarditis hub genes via integrated bioinformatics analysis. In addition, receiver operating characteristic curves constructed based on the expression levels of TBKBP1 and ERGIC1 could effectively distinguish healthy control individuals from patients with COVID-19. Functional enrichment analysis suggested several enriched biological pathways related to inflammation and immune response. Immune cell changes correlated with TBKBP1 and ERGIC1 levels in patients with COVID-19 or patients with COVID-19 and myocarditis. Tamibarotene, methotrexate and theophylline were identified as a potential drug targeting TBKBP1 and ERGIC1. ConclusionTBKBP1 and ERGIC1 were identified as crucial genes in the development of COVID-19-related myocarditis and have demonstrated a strong association with innate antiviral immunity. The present work may be helpful for further investigation of the molecular mechanisms and new therapeutic drug targets correlated with myocarditis in COVID-19.

Unveiling dysregulated lncRNAs and networks in non-syndromic cleft lip with or without cleft palate pathogenesis

Non-syndromic cleft lip with or without cleft palate (NSCL/P) is a common congenital facial malformation with a complex, incompletely understood origin. Long noncoding RNAs (lncRNAs) have emerged as pivotal regulators of gene expression, potentially shedding light on NSCL/P's etiology. This study aimed to identify critical lncRNAs and construct regulatory networks to unveil NSCL/P's underlying molecular mechanisms. Integrating gene expression profiles from the Gene Expression Omnibus (GEO) database, we pinpointed 30 dysregulated NSCL/P-associated lncRNAs. Subsequent analyses enabled the creation of competing endogenous RNA (ceRNA) networks, lncRNA-RNA binding protein (RBP) interaction networks, and lncRNA cis and trans regulation networks. RT-qPCR was used to examine the regulatory networks of lncRNA in vivo and in vitro. Furthermore, protein levels of lncRNA target genes were validated in human NSCL/P tissue samples and murine palatal shelves. Consequently, two lncRNAs and three mRNAs: FENDRR (log2FC = − 0.671, P = 0.040), TPT1-AS1 (log2FC = 0.854, P = 0.003), EIF3H (log2FC = − 1.081, P = 0.041), RBBP6 (log2FC = 0.914, P = 0.037), and SRSF1 (log2FC = 0.763, P = 0.026) emerged as potential contributors to NSCL/P pathogenesis. Functional enrichment analyses illuminated the biological functions and pathways associated with these lncRNA-related networks in NSCL/P. In summary, this study comprehensively delineates the dysregulated transcriptional landscape, identifies associated lncRNAs, and reveals pivotal sub-networks relevant to NSCL/P development, aiding our understanding of its molecular progression and setting the stage for further exploration of lncRNA and mRNA regulation in NSCL/P.

Constructing a competitive endogenous RNA network of EndMT-related atherosclerosis through weighted gene co-expression network analysis.

Atherosclerosis is a chronic inflammatory disease characterized by endothelial dysfunction and plaque formation. Under pro-inflammatory conditions, endothelial cells can undergo endothelial-to-mesenchymal transition (EndMT), contributing to atherosclerosis development. However, the specific regulatory mechanisms by which EndMT contributes to atherosclerosis remain unclear and require further investigation. Dan-Shen-Yin (DSY), a traditional Chinese herbal formula, is commonly used for cardiovascular diseases, but its molecular mechanisms remain elusive. Emerging evidence indicates that competing endogenous RNA (ceRNA) networks play critical roles in atherosclerosis pathogenesis. In this study, we constructed an EndMT-associated ceRNA network during atherosclerosis progression by integrating gene expression profiles from the Gene Expression Omnibus (GEO) database and weighted gene co-expression network analysis. Functional enrichment analysis revealed this EndMT-related ceRNA network is predominantly involved in inflammatory responses. ROC curve analysis showed the identified hub genes can effectively distinguish between normal vasculature and atherosclerotic lesions. Furthermore, Kaplan-Meier analysis demonstrated that high expression of IL1B significantly predicts ischemic events in atherosclerosis. Molecular docking revealed most DSY bioactive components can bind key EndMT-related lncRNAs, including AC003092.1, MIR181A1HG, MIR155HG, WEE2-AS1, and MIR137HG, suggesting DSY may mitigate EndMT in atherosclerosis by modulating the ceRNA network.

DRLM: A Robust Drug Representation Learning Method and its Applications.

Learning representations from data is a fundamental step for machine learning. High-quality and robust drug representations can broaden the understanding of pharmacology, and improve the modeling of multiple drug-related prediction tasks, which further facilitates drug development. Although there are a number of models developed for drug representation learning from various data sources, few researches extract drug representations from gene expression profiles. Since gene expression profiles of drug-treated cells are widely used in clinical diagnosis and therapy, it is believed that leveraging them to eliminate cell specificity can promote drug representation learning. In this paper, we propose a three-stage deep learning method for drug representation learning, named DRLM, which integrates gene expression profiles of drug-related cells and the therapeutic use information of drugs. Firstly, we construct a stacked autoencoder to learn low-dimensional compact drug representations. Secondly, we utilize an iterative clustering module to reduce the negative effects of cell specificity and noise in gene expression profiles on the low-dimensional drug representations. Thirdly, a therapeutic use discriminator is designed to incorporate therapeutic use information into the drug representations. The visualization analysis of drug representations demonstrates DRLM can reduce cell specificity and integrate therapeutic use information effectively. Extensive experiments on three types of prediction tasks are conducted based on different drug representations, and they show that the drug representations learned by DRLM outperform other representations in terms of most metrics. The ablation analysis also demonstrates DRLM's effectiveness of merging the gene expression profiles with the therapeutic use information.

Transcription and chromatin regulation by TAF4b during cellular quiescence of developing prospermatogonia.

Prospermatogonia (ProSpg) link the embryonic development of male primordial germ cells to the healthy establishment of postnatal spermatogonia and spermatogonial stem cells. While these spermatogenic precursor cells undergo the characteristic transitions of cycling and quiescence, the transcriptional events underlying these developmental hallmarks remain unknown. Here, we investigated the expression and function of TBP-associated factor 4b (Taf4b) in the timely development of quiescent mouse ProSpg using an integration of gene expression profiling and chromatin mapping. We find that Taf4b mRNA expression is elevated during the transition of mitotic-to-quiescent ProSpg and Taf4b-deficient ProSpg are delayed in their entry into quiescence. Gene ontology, protein network analysis, and chromatin mapping demonstrate that TAF4b is a direct and indirect regulator of chromatin and cell cycle-related gene expression programs during ProSpg quiescence. Further validation of these cell cycle mRNA changes due to the loss of TAF4b was accomplished via immunostaining for proliferating cell nuclear antigen (PCNA). Together, these data indicate that TAF4b is a key transcriptional regulator of the chromatin and quiescent state of the developing mammalian spermatogenic precursor lineage.

Integrated gene expression profiling and functional enrichment analyses to discover biomarkers and pathways associated with Guillain-Barré syndrome and autism spectrum disorder to identify new therapeutic targets

Guillain-Barré syndrome (GBS) is one of the most prominent and acute immune-mediated peripheral neuropathy, while autism spectrum disorders (ASD) are a group of heterogeneous neurodevelopmental disorders. The complete mechanism regarding the neuropathophysiology of these disorders is still ambiguous. Even after recent breakthroughs in molecular biology, the link between GBS and ASD remains a mystery. Therefore, we have implemented well-established bioinformatic techniques to identify potential biomarkers and drug candidates for GBS and ASD. 17 common differentially expressed genes (DEGs) were identified for these two disorders, which later guided the rest of the research. Common genes identified the protein-protein interaction (PPI) network and pathways associated with both disorders. Based on the PPI network, the constructed hub gene and module analysis network determined two common DEGs, namely CXCL9 and CXCL10, which are vital in predicting the top drug candidates. Furthermore, coregulatory networks of TF-gene and TF-miRNA were built to detect the regulatory biomolecules. Among drug candidates, imatinib had the highest docking and MM-GBSA score with the well-known chemokine receptor CXCR3 and remained stable during the 100 ns molecular dynamics simulation validated by the principal component analysis and the dynamic cross-correlation map. This study predicted the gene-based disease network for GBS and ASD and suggested prospective drug candidates. However, more in-depth research is required for clinical validation. Communicated by Ramaswamy H. Sarma

Spatial-MGCN: a novel multi-view graph convolutional network for identifying spatial domains with attention mechanism.

Recent advances in spatial transcriptomics technologies have enabled gene expression profiles while preserving spatial context. Accurately identifying spatial domains is crucial for downstream analysis and it requires the effective integration of gene expression profiles and spatial information. While increasingly computational methods have been developed for spatial domain detection, most of them cannot adaptively learn the complex relationship between gene expression and spatial information, leading to sub-optimal performance. To overcome these challenges, we propose a novel deep learning method named Spatial-MGCN for identifying spatial domains, which is a Multi-view Graph Convolutional Network (GCN) with attention mechanism. We first construct two neighbor graphs using gene expression profiles and spatial information, respectively. Then, a multi-view GCN encoder is designed to extract unique embeddings from both the feature and spatial graphs, as well as their shared embeddings by combining both graphs. Finally, a zero-inflated negative binomial decoder is used to reconstruct the original expression matrix by capturing the global probability distribution of gene expression profiles. Moreover, Spatial-MGCN incorporates a spatial regularization constraint into the features learning to preserve spatial neighbor information in an end-to-end manner. The experimental results show that Spatial-MGCN outperforms state-of-the-art methods consistently in several tasks, including spatial clustering and trajectory inference.

STAT3-CD163 cross-talk exhibits promising biomarkers for a progressive ischemic cardiomyopathy: integrative computational and gene expression profiling based on Gene Expression Omnibus datasets

Background and purpose: Ischemic heart disease frequently leads to heart failure, often resulting in death. In this study, we aimed to identify common hub mRNAs and pathways involved in the pathological progression of ischemic cardiomyopathy (ICM). Methods: Validation quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was carried out on peripheral blood and left ventricle specimens from patients in three groups with stable coronary artery disease (CAD), myocardial infarction (MI), and ICM and compared with corresponding controls. qRT-PCR was preceded by computational analysis of eight high-throughput RNA sequencing and microarray datasets from 499 patients and 233 controls, to determine possible common biologically meaningful differentially expressed genes (DEGs). To determine the potential pathological pathways, we performed Gene Ontology functional annotation, pathway enrichment analysis, protein–protein interaction (PPI) analysis, and constructed transcriptional factor/miRNA regulatory networks. Finally, approved drugs were screened. Results: Fifteen common DEGs with P < 0.01 were identified. STAT3, CEBPD, GLUL, and CD163 were hub-enriched mRNAs with an interaction score ≥ 0.50. Our qRT-PCR analysis showed an increased expression of STAT3 in all three patient groups and CD163, mainly in cardiac samples, in a remarkably ascending manner. Interaction modules showed co-regulators supporting high STAT3-CD163 connectivity, suggesting a potential role for STAT3-CD163 cross-talk-mediated inflammatory responses in ICM progression. Conclusions: Our results provided a novel perspective for understanding the underlying mechanisms of ICM progression and exploring new therapeutic agents. Clinical trial registration: URL: www.clinicaltrials.gov. Unique identifier: NCT05508269.

Integrated gene expression profiles reveal a transcriptomic network underlying the thermogenic response in adipose tissue

Obesity and type 2 diabetes are two closely related diseases representing a serious threat worldwide. An increase in metabolic rate through enhancement of non-shivering thermogenesis in adipose tissue may represent a potential therapeutic strategy. Nevertheless, a better understanding of thermogenesis transcriptional regulation is needed to allow the development of new effective treatments. Here, we aimed to characterize the specific transcriptomic response of white and brown adipose tissues after thermogenic induction. Using cold exposure to induce thermogenesis in mice, we identified mRNAs and miRNAs that were differentially expressed in several adipose depots. In addition, integration of transcriptomic data in regulatory networks of miRNAs and transcription factors allowed the identification of key nodes likely controlling metabolism and immune response. Moreover, we identified the putative role of the transcription factor PU.1 in the regulation of PPARγ-mediated thermogenic response of subcutaneous white adipose tissue. Therefore, the present study provides new insights into the molecular mechanisms that regulate non-shivering thermogenesis.

Novel mRNA Signature for Anti-TNF-α Therapy Primary Response in Patients With Ulcerative Colitis.

Ulcerative colitis (UC), an idiopathic, chronic inflammatory disorder of the colonic mucosa, is commonly treated with antitumor necrosis factor α (anti-TNF-α) agents. However, only approximately two-thirds have an initial response to these therapies. We integrated gene expression profiling from 3 independent data sets of 79 UC patients before they began anti-TNF-α therapy and calculated the differentially expressed genes between patient response and nonresponse to anti-TNF-α therapy and developed a de novo response-associated transcription signature score (logOR_Score) to demonstrate the predictive capability of anti-TNF-α therapy for therapeutic efficacy. Furthermore, we performed association analysis of the logOR_Score and clinical features, such as disease activity and immune microenvironment. A total of 2522 responsive and 1824 nonresponsive genes were identified from the integrated data set. Responsive genes were significantly enriched in metabolism-related pathways, whereas nonresponsive ones were associated with immune response-related pathways. The logOR_Score enabled the accurate prediction of the therapeutic efficacy of anti-TNF-α in 4 independent patient cohorts and outperformed the predictions made based on 6 transcriptome-based signatures. In terms of clinical features, the logOR_Score correlated highly with the activity of UC. From an immune microenvironment perspective, logOR_Scores of CD8+IL-17+ T cells, follicular B cells, and innate lymphoid cells significantly decreased in inflamed UC tissue. The de novo response-associated transcription signature may provide novel insights into the personalized treatment of patients with UC. Comprehensive analyses of the response-related subtypes and the association between logOR_Score and clinical features and immune microenvironment may provide insights into the underlying UC pathogenesis.

Fine mapping and cloning of a novel BrSCC1 gene for seed coat color in Brassica rapa L.

A novel BrSCC1 gene for seed coat color was fine mapped within a 41.1-kb interval on chromosome A03 in Brassica rapa and functionally validated by ectopic expression analysis. Yellow seed is a valuable breeding trait that can be potentiality applied for improving seed quality and oil productivity in oilseed Brassica crops. However, only few genes for yellow seed have been identified in B. rapa. We previously identified a minor quantitative trait locus (QTL), qSC3.1, for seed coat color on chromosome A03 in B. rapa. In order to isolate the seed coat color gene, a brown-seeded chromosome segment substitution line, CSSL-38, harboring the qSC3.1, was selected and crossed with the yellow-seeded recurrent parent, a rapid cycling inbred line of B. rapa (RcBr), to construct the secondary F2 population. Metabolite identification suggested that seed coat coloration in CSSL-38 was independent of proanthocyanidins (PAs) accumulation. Genetic analysis revealed that yellow seed was controlled by a single recessive gene, Seed Coat Color 1 (BrSCC1). Utilizing bulked segregant analysis (BSA)-seq and secondary F2 and F2:3 recombinants analysis, BrSCC1 was fine mapped within a 41.1-kb interval. By integrating gene expression profiling, genome sequence comparison, metabolite analysis, and functional validation through ectopic expression in Arabidopsis, the BraA03g040800.3C gene was confirmed to be BrSCC1, which positively correlated with the seed coat coloration. Our study provides a novel gene resource for the genetic improvement of yellow seeds in oilseed B. rapa.

A tensor decomposition-based integrated analysis applicable to multiple gene expression profiles without sample matching.

The integrated analysis of multiple gene expression profiles previously measured in distinct studies is problematic since missing both sample matches and common labels prevent their integration in fully data-driven, unsupervised training. In this study, we propose a strategy to enable the integration of multiple gene expression profiles among multiple independent studies with neither labeling nor sample matching using tensor decomposition unsupervised feature extraction. We apply this strategy to Alzheimer's disease (AD)-related gene expression profiles that lack precise correspondence among samples, including AD single-cell RNA sequence (scRNA-seq) data. We were able to select biologically reasonable genes using the integrated analysis. Overall, integrated gene expression profiles can function analogously to prior- and/or transfer-learning strategies in other machine-learning applications. For scRNA-seq, the proposed approach significantly reduces the required computational memory.

Drug repurposing using meta-analysis of gene expression in Alzheimer's disease.

Alzheimer's disease and other forms of dementia are disease that bring an increased global burden. However, the medicine developed to date remains limited. The purpose of this study is to predict drug repositioning candidates using a computational method that integrates gene expression profiles on Alzheimer's disease and compound-induced changes in gene expression levels. Gene expression data on Alzheimer's disease were obtained from the Gene Expression Omnibus (GEO) and we conducted a meta-analysis of their gene expression levels. The reverse scores of compound-induced gene expressions were computed based on the reversal relationship between disease and drug gene expression profiles. Reversal genes and the candidate compounds were identified by the leave-one-out cross-validation procedure. Additionally, the half-maximal inhibitory concentration (IC50) values and the blood-brain barrier (BBB) permeability of candidate compounds were obtained from ChEMBL and PubChem, respectively. New therapeutic target genes and drug candidates against Alzheimer's disease were identified by means of drug repositioning.

CEDA: integrating gene expression data with CRISPR-pooled screen data identifies essential genes with higher expression.

Clustered regularly interspaced short palindromic repeats (CRISPR)-based genetic perturbation screen is a powerful tool to probe gene function. However, experimental noises, especially for the lowly expressed genes, need to be accounted for to maintain proper control of false positive rate. We develop a statistical method, named CRISPR screen with Expression Data Analysis (CEDA), to integrate gene expression profiles and CRISPR screen data for identifying essential genes. CEDA stratifies genes based on expression level and adopts a three-component mixture model for the log-fold change of single-guide RNAs (sgRNAs). Empirical Bayesian prior and expectation-maximization algorithm are used for parameter estimation and false discovery rate inference. Taking advantage of gene expression data, CEDA identifies essential genes with higher expression. Compared to existing methods, CEDA shows comparable reliability but higher sensitivity in detecting essential genes with moderate sgRNA fold change. Therefore, using the same CRISPR data, CEDA generates an additional hit gene list. Supplementary data are available at Bioinformatics online.

Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef Cows.

Simple SummaryFemale reproductive failure is still a challenge for the beef industry. Several biological processes that underlie fertility-related traits, such as the establishment of pregnancy and embryo survival, are still unclear. Increased availability of transcriptomic data has allowed a deep investigation of the potential mechanisms involved in fertility. This study investigated candidate gene biomarkers predictive of pregnancy status and underlying fertility-related networks. To this end, we integrated gene expression profiles through supervised machine learning and gene network modeling. We identified nine biologically relevant endometrial gene biomarkers that could discriminate against pregnancy status in cows. These biomarkers were co-expressed with genes critical for uterine receptivity, including endometrial tissue remodeling, focal adhesion, and embryo development. This study outlined key pathways involved with pregnancy success and provided predictive candidate biomarkers for pregnancy outcome in cows.Reproductive failure is still a challenge for beef producers and a significant cause of economic loss. The increased availability of transcriptomic data has shed light on the mechanisms modulating pregnancy success. Furthermore, new analytical tools, such as machine learning (ML), provide opportunities for data mining and uncovering new biological events that explain or predict reproductive outcomes. Herein, we identified potential biomarkers underlying pregnancy status and fertility-related networks by integrating gene expression profiles through ML and gene network modeling. We used public transcriptomic data from uterine luminal epithelial cells of cows retrospectively classified as pregnant (P, n = 25) and non-pregnant (NP, n = 18). First, we used a feature selection function from BioDiscML and identified SERPINE3, PDCD1, FNDC1, MRTFA, ARHGEF7, MEF2B, NAA16, ENSBTAG00000019474, and ENSBTAG00000054585 as candidate biomarker predictors of pregnancy status. Then, based on co-expression networks, we identified seven genes significantly rewired (gaining or losing connections) between the P and NP networks. These biomarkers were co-expressed with genes critical for uterine receptivity, including endometrial tissue remodeling, focal adhesion, and embryo development. We provided insights into the regulatory networks of fertility-related processes and demonstrated the potential of combining different analytical tools to prioritize candidate genes.