Transcriptome Expression Data Research Articles

Multimodal AI/ML for discovering novel biomarkers and predicting disease using multi-omics profiles of patients with cardiovascular diseases

Cardiovascular diseases (CVDs) are complex, multifactorial conditions that require personalized assessment and treatment. Advancements in multi-omics technologies, namely RNA sequencing and whole-genome sequencing, have provided translational researchers with a comprehensive view of the human genome. The efficient synthesis and analysis of this data through integrated approach that characterizes genetic variants alongside expression patterns linked to emerging phenotypes, can reveal novel biomarkers and enable the segmentation of patient populations based on personalized risk factors. In this study, we present a cutting-edge methodology rooted in the integration of traditional bioinformatics, classical statistics, and multimodal machine learning techniques. Our approach has the potential to uncover the intricate mechanisms underlying CVD, enabling patient-specific risk and response profiling. We sourced transcriptomic expression data and single nucleotide polymorphisms (SNPs) from both CVD patients and healthy controls. By integrating these multi-omics datasets with clinical demographic information, we generated patient-specific profiles. Utilizing a robust feature selection approach, we identified a signature of 27 transcriptomic features and SNPs that are effective predictors of CVD. Differential expression analysis, combined with minimum redundancy maximum relevance feature selection, highlighted biomarkers that explain the disease phenotype. This approach prioritizes both biological relevance and efficiency in machine learning. We employed Combination Annotation Dependent Depletion scores and allele frequencies to identify variants with pathogenic characteristics in CVD patients. Classification models trained on this signature demonstrated high-accuracy predictions for CVD. The best performing of these models was an XGBoost classifier optimized via Bayesian hyperparameter tuning, which was able to correctly classify all patients in our test dataset. Using SHapley Additive exPlanations, we created risk assessments for patients, offering further contextualization of these predictions in a clinical setting. Across the cohort, RPL36AP37 and HBA1 were scored as the most important biomarkers for predicting CVDs. A comprehensive literature review revealed that a substantial portion of the diagnostic biomarkers identified have previously been associated with CVD. The framework we propose in this study is unbiased and generalizable to other diseases and disorders.

Breast cancer promotes the expression of neurotransmitter receptor related gene groups and image simulation of prognosis model

Breast cancer (BC), a prevalent and severe malignancy, detrimentally affects women globally. Its prognostic implications are profoundly influenced by gene expression patterns. This study retrieved 509 BC-associated oncogenes and 1,012 neurotransmitter receptor-related genes from the GSEA and KEGG databases, intersecting to identify 98 relevant genes. Clinical and transcriptomic expression data related to BC were downloaded from the TCGA, and differential genes were identified based on an FDR value <0.05 & |log2FC| ≥ 0.585. Univariate analysis of these genes revealed that high expression of NSF and low expression of HRAS, KIF17, and RPS6KA1 are closely associated with BC survival prognosis. A prognostic model constructed for these four genes demonstrated significant prognostic relevance for BC-TCGA patients (P < 0.001). Subsequently, an immunofunctional analysis of the BC oncogene-neurotransmitter receptor-related gene cluster revealed the involvement of immune cells such as T cells CD8, T cells CD4 memory resting, and Macrophages M2. Further analysis indicated that immune functions were primarily concentrated in APC_co_inhibition, APC_co_stimulation, CCR, and Check-point, among others. Lastly, a prognostic nomogram model was established, and ROC curve analysis revealed that the nomogram is a vital indicator for assessing BC prognosis, with 1-year, 3-year, and 5-year survival rates of 0.981, 0.897, and 0.802, respectively. This model demonstrates high calibration, clinical utility, and predictive capability, promising to offer an effective preliminary tool for clinical diagnostics.

Increased UBD Is a Potential Diagnostic and Prognostic Biomarker in Glioma.

Various studies have demonstrated that ubiquitin D (UBD) is overexpressed in different cancer types and may serve as a potential prognostic factor. However, additional research is necessary to establish the prognostic significance and possible role of UBD in glioma. Transcriptomic expression data from The Cancer Genome Atlas database (TCGA) and Chinese Glioma Genome Atlas (CGGA) were analyzed to identify UBD expression differences in tumor and normal tissues. The relative levels of UBD in glioma and normal tissues were determined using qRT-PCR and WB. Logistic regression analysis was performed to investigate the association between UBD expression and clinicopathological characteristics of glioma patients. To evaluate the diagnostic and prognostic predictive values of UBD, we used Kaplan-Meier survival curves, Cox regression analysis, diagnostic receiver operating characteristic (ROC) curves, and nomogram model. We also conducted wound healing assays, transwell assays, EdU assays, and colony formation assays to verify the UBD function. Gene ontology (GO) enrichment analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, as well as gene set enrichment analysis (GSEA), were employed to determine the functions of UBD. Finally, we performed the western blot assays to assess changes in EMT markers as well as p-PI3K, p-AKT, and p-mTOR expressions. Our study revealed a remarkable increase of UBD expression in glioma samples. Cox regression analysis demonstrated that high expression of UBD mRNA was an independent prognostic factor for overall survival (OS) in TCGA. ROC curve analysis showed that UBD expression levels could differentiate glioma from adjacent normal tissues accurately. Additionally, knockdown of UBD reduced the migration, invasion, and proliferation ability of glioma cells while UBD overexpression had the opposite effect. GSEA showed that the expression of UBD involved with various pathways including epithelial-mesenchymal transition (EMT), PI3K-AKT-mTOR signaling, P53 pathway, angiogenesis, inflammatory response, KRAS signaling, hypoxia, as well as TGF-β signaling. Furthermore, our findings suggest that UBD accelerates the activation of EMT and PI3K/AKT/mTOR pathway.

Infiltrating B-cell subtypes and associated hub genes in nasopharyngeal carcinoma identified from integrated single-cell and bulk-RNA sequencing data.

135 Background: Nasopharyngeal carcinoma (NPC) is associated with lymphocyte infiltration; however, the majority of research on NPC has focused on the role of T cells, with relatively little known about the roles of B cells and their subtypes. Therefore, we evaluated the prognostic value of CD20+ B cell density and B-cell subtypes along with their functional enrichment and hub genes in NPC. Methods: The prognostic value of CD20+ B-cell density for distant metastasis-free survival (DMFS), overall survival (OS), and progression-free survival (PFS) was explored by immunohistochemistry using multivariate analysis. Transcriptomic expression data from Gene Expression Omnibus (GEO) datasets were analyzed to identify B-cell subtypes and their functional enrichment in NPC tissues. Pseudotime trajectory analysis was performed to evaluate the B-cell differentiation trajectory and hub genes were identified using Cytoscape software. Results: Patients with NPC exhibiting a high infiltrating density of CD20+ B cells showed significantly better 5-year DMFS, OS, and PFS compared to those of patients with a low infiltrating density. Naïve B cells, switched memory B cells, exhausted B cells, and plasma cells were identified as key B-cell subtypes infiltrating NPC tumors, with naïve B cells showing the highest infiltration levels associated with a better prognosis. Naïve B cells were closely associated with immune pathways and the hub genes were typical markers for T and B cells. Conclusions: A high infiltrating density of B cells showed strong prognostic value in patients with NPC. Naïve B cells may play an important role in tumor immunity for NPC.

Genome-Wide Association and RNA-Seq Analyses Reveal a Potential Candidate Gene Related to Oil Content in Soybean Seeds.

Soybean is a crucial crop globally, serving as a significant source of unsaturated fatty acids and protein in the human diet. However, further enhancements are required for the related genes that regulate soybean oil synthesis. In this study, 155 soybean germplasms were cultivated under three different environmental conditions, followed by phenotypic identification and genome-wide association analysis using simplified sequencing data. Genome-wide association analysis was performed using SLAF-seq data. A total of 36 QTLs were significantly associated with oil content (-log10(p) > 3). Out of the 36 QTLs associated with oil content, 27 exhibited genetic overlap with previously reported QTLs related to oil traits. Further transcriptome sequencing was performed on extreme high-low oil soybean varieties. Combined with transcriptome expression data, 22 candidate genes were identified (|log2FC| ≥ 3). Further haplotype analysis of the potential candidate genes showed that three potential candidate genes had excellent haplotypes, including Glyma.03G186200, Glyma.09G099500, and Glyma.18G248900. The identified loci harboring beneficial alleles and candidate genes likely contribute significantly to the molecular network's underlying marker-assisted selection (MAS) and oil content.

Molecular Cloning and Functional Identification of a Pericarp- and Testa-Abundant Gene's (AhN8DT-2) Promoter from Arachis hypogaea.

Cultivated peanut (Arachis hypogaea L.) is a key oil- and protein-providing legume crop of the world. It is full of nutrients, and its nutrient profile is comparable to that of other nuts. Peanut is a unique plant as it showcases a pegging phenomenon, producing flowers above ground, and after fertilization, the developing peg enters the soil and produces seeds underground. This geocarpic nature of peanut exposes its seeds to soil pathogens. Peanut seeds are protected by an inedible pericarp and testa. The pericarp- and testa-specific promoters can be effectively used to improve the seed defense. We identified a pericarp- and testa-abundant expression gene (AhN8DT-2) from available transcriptome expression data, whose tissue-specific expression was further confirmed by the qRT-PCR. The 1827bp promoter sequence was used to construct the expression vector using the pMDC164 vector for further analysis. Quantitative expression of the GUS gene in transgenic Arabidopsis plants showed its high expression in the pericarp. GUS staining showed a deep blue color in the pericarp and testa. Cryostat sectioning of stained Arabidopsis seeds showed that expression is only limited to seed coat (testa), and staining was not present in cotyledons and embryos. GUS staining was not detected in any other tissues, including seedlings, leaves, stems, and roots, except for some staining in flowers. Under different phytohormones, this promoter did not show an increase in expression level. These results indicated that the AhN8DT-2 promoter drives GUS gene expression in a pericarp- and testa-specific manner. The identified promoter can be utilized to drive disease resistance genes, specifically in the pericarp and testa, enhancing peanut seed defense against soil-borne pathogens. This approach has broader implications for improving the resilience of peanut crops and other legumes, contributing to sustainable agricultural practices and food security.

Ubiquitination-Related Gene Signature, Nomogram and Immune Features for Prognostic Prediction in Patients with Head and Neck Squamous Cell Carcinoma.

The objective of this research was to create a prognostic model focused on genes related to ubiquitination (UbRGs) for evaluating their clinical significance in head and neck squamous cell carcinoma (HNSCC) patients. The transcriptome expression data of UbRGs were obtained from The Cancer Genome Atlas (TCGA) database, and weighted gene co-expression network analysis (WGCNA) was used to identify specific UbRGs within survival-related hub modules. A multi-gene signature was formulated using LASSO Cox regression analysis. Furthermore, various analyses, including time-related receiver operating characteristics (ROCs), Kaplan-Meier, Cox regression, nomogram prediction, gene set enrichment, co-expression, immune, tumor mutation burden (TMB), and drug sensitivity, were conducted. Ultimately, a prognostic signature consisting of 11 gene pairs for HNSCC was established. The Kaplan-Meier curves indicated significantly improved overall survival (OS) in the low-risk group compared to the high-risk group (p < 0.001), suggesting its potential as an independent and dependable prognostic factor. Additionally, a nomogram with AUC values of 0.744, 0.852, and 0.861 at 1-, 3-, and 5-year intervals was developed. Infiltration of M2 macrophages was higher in the high-risk group, and the TMB was notably elevated compared to the low-risk group. Several chemotherapy drugs targeting UbRGs were recommended for low-risk and high-risk patients, respectively. The prognostic signature derived from UbRGs can effectively predict prognosis and provide new personalized therapeutic targets for HNSCC.

P-707 Discovery of novel serum biomarkers for premature ovarian insufficiency through machine-learning bioinformatic analysis

Abstract Study question Can we develop predictive, practical biomarkers for premature ovarian insufficiency (POI) from normal ovaries and human blood, tested and validated by machine-learning (ML) algorithm? Summary answer Applying random forest and XGBoost algorithms, we pinpointed 60 significant genes from ovarian tissues; transcriptome analysis identified RPM1 as the most significant biomarker. What is known already Accelerated ovarian aging has been suggested as one of the possible underlying mechanisms of POI; previous literature has focused on identifying its cause in genetic realm, commonly adopting first-generation sequencing (Sanger) and next-generation sequencing (NGS). However, such powerful tools are still limited in accurately depicting the complete transcript information of ovarian tissue. Moreover, theoretical value of such disease-predictive gene candidates needs to be validated using human tissues and blood, and finally, underlying mechanisms should be confirmed with ideal animal model and cell line. Study design, size, duration To overcome the existing limitations and maximize clinical relevance of disease-predictive gene analysis and theoretical biomarker exploration, the current study incorporated the transcriptome expression data from both young and old ovaries obtained from the Genotype-Tissue Expression (GTEx) database. Using supervised machine learning techniques, potential marker genes were identified, and a meaningful predictive index was subsequently developed using human plasma samples of control and POI women with and without pregnancy history. Participants/materials, setting, methods The performance evaluation of two ML models were achieved using the area under receiver operating characteristic curve (AUROC), accuracy, kappa, and F1 score, among various age groups. The expression profiles of final candidate marker genes were analyzed within the GTEx datasets, in relation to their correlation pattern with age. Finally, the potential transcriptome biomarkers were quantitatively evaluated using human blood samples from 92 controls and 108 POI women visiting a tertiary center in 2023. Main results and the role of chance We found 41 significant genes using the random forest algorithm and 19 significant genes using the XGboost algorithm. We then tested the ML-characterized 7 final candidate biomarkers in human blood samples (all p &amp;lt; 0.05). The AUROC and interventional predictive index results demonstrated the potential of RPM1 as a predictive marker with AUROC of 0.784, comparable to the AUROC of anti-mullerian hormone (AMH) as 0.659 and inhibin B (INHB) as 0.646. Limitations, reasons for caution To avoid potential biases, RPMs should be tested in a larger, multinational cohort with different genetic and geographic backgrounds, possibly incorporating Mendelian randomization method to estimate the causal effect of RPMs on POI. Finally, to ultimately confirm their therapeutic potential, preclinical studies should be considered using RPM recombinant proteins. Wider implications of the findings Our results proposed that predictive biomarkers for POI could be associated with premature, accelerated ovarian aging of the disease mechanism, and these biomarkers can be developed as evaluation metrics for the efficacy, progress, and prognosis of interventions and treatment modalities assigned to POI women seeking fertility. Trial registration number 202302240001

Time-course transcriptome data of silk glands in day 0–7 last-instar larvae of Bombyx mori (w1 pnd strain)

Time-course transcriptome expression data were constructed for four parts of the silk gland (anterior, middle, and posterior parts of the middle silk gland, along with the posterior silk gland) in the domestic silkworm, Bombyx mori, from days 0 to 7 of the last-instar larvae. For sample preparation, silk glands were extracted from one female and one male larva every 24 hours accurately after the fourth ecdysis. The reliability of these transcriptome data was confirmed by comparing the transcripts per million (TPM) values of the silk gene and quantitative reverse transcription PCR results. Hierarchical cluster analysis results supported the reliability of transcriptome data. These data are likely to contribute to the progress in molecular biology and genetic research using B. mori, such as elucidating the mechanism underlying the massive production of silk proteins, conducting entomological research using a meta-analysis as a model for lepidopteran insect species, and exploring medical research using B. mori as a model for disease species by utilising transcriptome data.

Combinatorial metabolomic and transcriptomic analysis of muscle growth in hybrid striped bass (female white bass Morone chrysops x male striped bass M. saxatilis)

BackgroundUnderstanding growth regulatory pathways is important in aquaculture, fisheries, and vertebrate physiology generally. Machine learning pattern recognition and sensitivity analysis were employed to examine metabolomic small molecule profiles and transcriptomic gene expression data generated from liver and white skeletal muscle of hybrid striped bass (white bass Morone chrysops x striped bass M. saxatilis) representative of the top and bottom 10 % by body size of a production cohort.ResultsLarger fish (good-growth) had significantly greater weight, total length, hepatosomatic index, and specific growth rate compared to smaller fish (poor-growth) and also had significantly more muscle fibers of smaller diameter (≤ 20 µm diameter), indicating active hyperplasia. Differences in metabolomic pathways included enhanced energetics (glycolysis, citric acid cycle) and amino acid metabolism in good-growth fish, and enhanced stress, muscle inflammation (cortisol, eicosanoids) and dysfunctional liver cholesterol metabolism in poor-growth fish. The majority of gene transcripts identified as differentially expressed between groups were down-regulated in good-growth fish. Several molecules associated with important growth-regulatory pathways were up-regulated in muscle of fish that grew poorly: growth factors including agt and agtr2 (angiotensins), nicotinic acid (which stimulates growth hormone production), gadd45b, rgl1, zfp36, cebpb, and hmgb1; insulin-like growth factor signaling (igfbp1 and igf1); cytokine signaling (socs3, cxcr4); cell signaling (rgs13, rundc3a), and differentiation (rhou, mmp17, cd22, msi1); mitochondrial uncoupling proteins (ucp3, ucp2); and regulators of lipid metabolism (apoa1, ldlr). Growth factors pttg1, egfr, myc, notch1, and sirt1 were notably up-regulated in muscle of good-growing fish.ConclusionA combinatorial pathway analysis using metabolomic and transcriptomic data collectively suggested promotion of cell signaling, proliferation, and differentiation in muscle of good-growth fish, whereas muscle inflammation and apoptosis was observed in poor-growth fish, along with elevated cortisol (an anti-inflammatory hormone), perhaps related to muscle wasting, hypertrophy, and inferior growth. These findings provide important biomarkers and mechanisms by which growth is regulated in fishes and other vertebrates as well.

Oxidative DNA damage contributes to usnic acid-induced toxicity in human induced pluripotent stem cell-derived hepatocytes.

Dietary supplements containing usnic acid have been increasingly marketed for weight loss over the past decades, even though incidences of severe hepatotoxicity and acute liver failure due to their overuse have been reported. To date, the toxic mechanism of usnic acid-induced liver injury at the molecular level still remains to be fully elucidated. Here, we conducted a transcriptomic study on usnic acid using a novel in vitro hepatotoxicity model employing human induced pluripotent stem cell (iPSC)-derived hepatocytes. Treatment with 20 μM usnic acid for 24 h caused 4272 differentially expressed genes (DEGs) in the cells. Ingenuity Pathway Analysis (IPA) based on the DEGs and gene set enrichment analysis (GSEA) using the whole transcriptome expression data concordantly revealed several signaling pathways and biological processes that, when taken together, suggest that usnic acid caused oxidative stress and DNA damage in the cells, which further led to cell cycle arrest and eventually resulted in cell death through apoptosis. These transcriptomic findings were subsequently corroborated by a variety of cellular assays, including reactive oxygen species (ROS) generation and glutathione (GSH) depletion, DNA damage (pH2AX detection and 8-hydroxy-2'-deoxyguanosine [8-OH-dg] assay), cell cycle analysis, and caspase 3/7 activity. Collectively, the results of the current study accord with previous in vivo and in vitro findings, provide further evidence that oxidative stress-caused DNA damage contributes to usnic acid-induced hepatotoxicity, shed new light on molecular mechanisms of usnic acid-induced hepatotoxicity, and demonstrate the usefulness of iPSC-derived hepatocytes as an in vitro model for hepatotoxicity testing and prediction.

Comprehensive analysis of immune-related lncRNAs in AML patients uncovers potential therapeutic targets and prognostic biomarkers

PurposeThe objective of this study was to provide theoretically feasible strategies by understanding the relationship between the immune microenvironment and the diagnosis and prognosis of AML patients. To this end, we built a ceRNA network with lncRNAs as the core and analyzed the related lncRNAs in the immune microenvironment by bioinformatics analysis. MethodsAML transcriptome expression data and immune-related gene sets were obtained from TCGA and ImmPort. Utilizing Pearson correlation analysis, differentially expressed immune-related lncRNAs were identified. Then, the LASSO-Cox regression analysis was performed to generate a risk signature consisting immune-related lncRNAs. Accuracy of signature in predicting patient survival was evaluated using univariate and multivariate analysis. Next, GO and KEGG gene enrichment and ssGSEA were carried out for pathway enrichment analysis of 183 differentially expressed genes, followed by drug sensitivity and immune infiltration analysis with pRRophetic and CIBERSORT, respectively. Cytoscape was used to construct the ceRNA network for these lncRNAs. Results816 common lncRNAs were selected to acquire the components related to prognosis. The final risk signature established by multivariate Cox and stepwise regression analysis contained 12 lncRNAs engaged in tumor apoptotic and metastatic processes: LINC02595, HCP5, AC020934.2, AC008770.3, LINC01770, AC092718.4, AL589863.1, AC131097.4, AC012368.1, C1RL-AS1, STARD4-AS1, and AC243960.1. Based on this predictive model, high-risk patients exhibited lower overall survival rates than low-risk patients. Signature lncRNAs showed significant correlation with tumor-infiltrating immune cells. In addition, significant differences in PD-1/PD-L1 expression and bleomycin/paclitaxel sensitivity were observed between risk groups. ConclusionLncRNAs related to immune microenvironment were prospective prognostic and therapeutic options for AML.

Molecular mechanisms underlying structural plasticity of electroconvulsive therapy in major depressive disorder.

Although previous studies reported structural changes associated with electroconvulsive therapy (ECT) in major depressive disorder (MDD), the underlying molecular basis of ECT remains largely unknown. Here, we combined two independent structural MRI datasets of MDD patients receiving ECT and transcriptomic gene expression data from Allen Human Brain Atlas to reveal the molecular basis of ECT for MDD. We performed partial least square regression to explore whether/how gray matter volume (GMV) alterations were associated with gene expression level. Functional enrichment analysis was conducted using Metascape to explore ontological pathways of the associated genes. Finally, these genes were further assigned to seven cell types to determine which cell types contribute most to the structural changes in MDD patients after ECT. We found significantly increased GMV in bilateral hippocampus in MDD patients after ECT. Transcriptome-neuroimaging association analyses showed that expression levels of 726 genes were positively correlated with the increased GMV in MDD after ECT. These genes were mainly involved in synaptic signaling, calcium ion binding and cell-cell signaling, and mostly belonged to excitatory and inhibitory neurons. Moreover, we found that the MDD risk genes of CNR1, HTR1A, MAOA, PDE1A, and SST as well as ECT related genes of BDNF, DRD2, APOE, P2RX7, and TBC1D14 showed significantly positive associations with increased GMV. Overall, our findings provide biological and molecular mechanisms underlying structural plasticity induced by ECT in MDD and the identified genes may facilitate future therapy for MDD.

Integrated analysis of transcriptome and miRNAome reveals the heat stress response of Pinellia ternata seedlings

Pinellia ternata (Thunb.) Briet., a valuable herb native to China, is susceptible to the “sprout tumble” phenomenon because of high temperatures, resulting in a significant yield reduction. However, the molecular regulatory mechanisms underlying the response of P. ternata to heat stress are not well understood. In this study, we integrated transcriptome and miRNAome sequencing to identify heat-response genes, microRNAs (miRNAs), and key miRNA-target pairs in P. ternata that differed between heat-stress and room-temperature conditions. Transcriptome analysis revealed extensive reprogramming of 4,960 genes across various categories, predominantly associated with cellular and metabolic processes, responses to stimuli, biological regulation, cell parts, organelles, membranes, and catalytic and binding activities. miRNAome sequencing identified 1,597 known/conserved miRNAs that were differentially expressed between the two test conditions. According to the analysis, genes and miRNAs associated with the regulation of transcription, DNA template, transcription factor activity, and sequence-specific DNA binding pathways may play a major role in the resistance to heat stress in P. ternata. Integrated analysis of the transcriptome and miRNAome expression data revealed 41 high-confidence miRNA-mRNA pairs, forming 25 modules. MYB-like proteins and calcium-responsive transcription coactivators may play an integral role in heat-stress resistance in P. ternata. Additionally, the candidate genes and miRNAs were subjected to quantitative real-time polymerase chain reaction to validate their expression patterns. These results offer a foundation for future studies exploring the mechanisms and critical genes involved in heat-stress resistance in P. ternata.

ShinyExprPortal: a configurable 'shiny' portal for sharing analysis of molecular expression data.

The scale of omics research presents many obstacles to full sharing and access to analysis results. Current publication models impose limits on the number of pages and figures, requiring careful preparation and selection of content. At the same time, depositing data in open repositories significantly shifts the burden of access and reproduction to readers, who may include people who are not programmers or analysts. We introduce shinyExprPortal, an R package that implements omics web portals with minimal coding effort. The portals allow exploration of transcriptomic or proteomic expression data and phenotypes, showcasing results of various types of analysis including differential expression, co-expression and pathways analysis. The integration with bioinformatics workflows enables researchers to focus on their results and share findings using interactive and publication-quality plots. The shinyExprPortal package is available to download and install from CRAN and https://github.com/C4TB/shinyExprPortal.

Abstract 888: Uncovering clinically relevant omics signatures from pan-cancer imaging and multi-omics data integration

Abstract Despite advancements in deep learning for histopathology, integrating these insights with multi-omics data to uncover clinically relevant omics pathway-level signatures remains a challenge. Our study addresses this gap by applying unsupervised learning techniques on pan-cancer multi-omics data, leveraging 3,080 Hematoxylin and Eosin (H&amp;E) images from 1,010 patients in Clinical Proteomic Tumor Analysis Consortium (CPTAC) to uncover omics pathway-level signatures that drive discernable morphology phenotypes at the tissue level. First, imaging models were trained to predict clinical and mutation outcomes, and thereafter, integrated with transcriptomic and proteomic expression data using sparse canonical correlation analysis. Our findings reveal that images of TP53 mutated samples exhibited increased nuclear size and dense lymphoplasmacytic infiltration. These morphological changes correlated with neutrophil and macrophage signaling at the proteomic level, and IL-1 mediated signaling at the transcriptomic level, highlighting the complementary perspectives different omics can provide. To further elucidate immune interactions, we applied multi-omics deconvolution to identify 7 immune subtypes, characterizing each with gene set enrichment, germline DNA variations, and kinase activations. We show that imaging models trained to predict these subtypes differentiate tissue morphologies corresponding to immune enrichment (AUROC: 0.84). To further confirm the robustness of our approach, we trained models to predict co-regulated proteomic modules clustered by independent component analysis. We demonstrate a signature in pan-squamous tumors, consisting of T cell markers and interferon-gamma response proteins like CD4, CD48, and GBP5, which imaging models can successfully predict from histopathology. Manual review from pathologists confirmed lymphocytic T-cell density as a differentiator in samples with the highest and lowest T-cell signaling. Together, these approaches demonstrate that genomic characteristics discovered via unbiased mining of pan-cancer multi-omics data manifest as quantitative imaging phenotypes. These results underscore the potential of multi-omics and digital pathology to integratively uncover and confirm new cancer biology. Our ongoing work will explore predicting drug response and survival from morphology patterns related to multi-omics signatures. Citation Format: Joshua Wang, Runyu Hong, Jimin Tan, Wenke Liu, David Fenyȯ. Uncovering clinically relevant omics signatures from pan-cancer imaging and multi-omics data integration [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 888.

Molecular insights into gastric cancer: The impact of TGFBR2 and hsa-mir-107 revealed by microarray sequencing and bioinformatics

BackgroundGastric carcinoma (GC) remains a significant therapeutic challenge, garnering widespread attention. Oxymatrine (OMT), an active component of the traditional Chinese medicine compound Kushen injection (CKI), has shown promising results in combination with chemotherapy for the treatment of GC. However, the molecular mechanisms underlying OMT's therapeutic effects in GC have yet to be elucidated. MethodsThe transcriptomic expression data of HGC-27 post-OMT intervention were obtained through microarray sequencing, while the miRNA and mRNA sequencing data for GC patients were sourced from the TCGA database. The mechanism of OMT intervention in GC is analyzed in multiple aspects, including Protein-Protein Interactions (PPI), Competitive Endogenous RNA (ceRNA) networks, correlation and co-expression analyses, immune infiltration, and clinical implications. ResultsBy analyzing key modules, five critical mRNAs were identified, and their interacting miRNAs were predicted to construct a ceRNA network. Among these, TGFBR2 and hsa-miR-107 have correlations or co-expression relationships with other genes in the network. They are differentially expressed in most other cancers, associated with prognosis, and have diagnostic value. TGFBR2 also exhibits immune infiltration phenomena, and its high expression is linked to poor patient prognosis. Low expression of hsa-miR-107 is associated with poor patient prognosis. OMT may act on the TGFβ/Smad signaling pathway or negatively regulate the WNT signaling pathway through the hsa-miR-107/BTRC axis, thereby inhibiting the onset and progression of GC. ConclusionThe mechanisms of OMT intervention in GC are diverse, TGFBR2 and hsa-miR-107 may serve as prognostic molecular biomarkers or potential therapeutic targets.

Dysfunction of CCT3-associated network signals for the critical state during progression of hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is one of the most common malignant tumors and is a serious threat to human health; thus, early diagnosis and adequate treatment are essential. However, there are still great challenges in identifying the tipping point and detecting early warning signals of early HCC. In this study, we aimed to identify the tipping point (critical state) of and key molecules involved in hepatocarcinogenesis based on time series transcriptome expression data of HCC patients. The phase from veHCC (very early HCC) to eHCC (early HCC) was identified as the critical state in HCC progression, with 143 genes identified as key candidate molecules by combining the DDRTree (dimensionality reduction via graph structure learning) and DNB (dynamic network biomarker) methods. Then, we ranked the candidate genes to verify their mRNA levels using the diethylnitrosamine (DEN)-induced HCC mouse model and identified five early warning signals, namely, CCT3, DSTYK, EIF3E, IARS2 and TXNRD1; these signals can be regarded as the potential early warning signals for the critical state of HCC. We identified CCT3 as an independent prognostic factor for HCC, and functions of CCT3 involving in the “MYCtargets_V1” and “E2F-Targets” are closely related to the progression of HCC. The predictive method combining the DDRTree and DNB methods can not only identify the key critical state before cancer but also determine candidate molecules of critical state, thus providing new insight into the early diagnosis and preemptive treatment of HCC.

Vasculogenic mimicry regulates immune infiltration and mutational status of the tumor microenvironment in breast cancer to influence tumor prognosis.

Vasculogenic mimicry (VM) refers to the direct formation of microcirculatory ducts by invasive malignant tumors via cellular phenotypic transformation. However, there is a lack of VM-based biomarkers for breast cancer. We obtained transcriptomic expression data, single cell sequencing data, and clinical data of patients from The Cancer Genome Atlas Program (TCGA) database and GEO database, performed single cell analysis to obtain specific type annotations of breast cancer cells and analyzed their spatial expression analysis. Kyoto Encyclopedia of Genes and Genomes (KEGG) and gene ontology (GO) analyses as well as Gene Set Enrichment Analysis (GSEA) analyses were performed to clarify the biological pathways and tumor functional enrichment relationships of the major expressed genes of VM in the breast cancer bulk data specimens. VM biomarkers were constructed. Meanwhile, the relationship between VM scores and tumor immune infiltration in breast cancer was analyzed using MCPcounter and ssGSEA methods. In addition, we assessed the specific relationship between NDRG1, a key VM gene in breast cancer, and tumor colonization, adhesion and invasion by biological experiments in breast cancer cell lines. The main cell types of breast cancer (BRCA) samples were annotated by single cell transcriptome analysis. Most of the VM-high group was present in epithelial cells, whereas the VM-low group was present in immune and stromal cells. Multiple tumor pathways such as TGFβ p53 and MAPK were closely associated with VM-mediated breast cancer infiltration and invasion. A prognostic model of breast cancer based on VM key genes was constituted. Prognostic stratification of breast cancer was successfully achieved for the TCGA-BRCA and GSE58812 datasets. Through immune infiltration analysis, we found that differential expression of VM markers was associated with multiple immune cell regulation. In MDA-MB-231 and MDA-MB-453 cell lines, we found that the NDRG1 gene significantly promoted colony formation of breast cancer cells. Our constructed VM-related gene-based model of breast cancer biology holds promise for prognostic prediction and patient stratification of breast cancer. This may provide a potentially clinically valuable aid in promoting a deeper understanding of the biological regulation of VM in breast cancer and exploring the specific mechanisms of tumor angiogenesis and breast cancer development.

Abstract A003: Global and TNBC-specific biomarkers of sensitivity and resistance in early stage breast cancer from the I-SPY2 biomarker data compendium

Abstract Introduction Diverse qualifying and exploratory sensitivity and resistance biomarkers were previously identified in the I-SPY2 trial. Notably, for TNBC patients, fewer predictive biomarkers were found compared with other receptor subgroups. Here we summarize and synthesize observations of biomarkers associated with sensitivity or resistance in the overall trial population and specifically in the TNBC population, to help further elucidate underlying biological mechanisms. Methods The I-SPY2 990 Data Compendium (PMID 35623341) contains pre-treatment full transcriptome expression data of 987 patients and quantitative expression data from 139 proteins/phosphoproteins from 736 patients across 10 treatment arms of the I-SPY2 trial. In this analysis, we re-organized and synthesized the pathologic complete response (pCR)-association results from previous analyses of these data to identify areas of overlap and discrepancies between global predictive biomarkers and TNBC-specific biomarkers, in all treatment arms together and in the control (chemotherapy) arm. Global biomarkers are defined as significantly associated with sensitivity or resistance in the overall trial population (all treatment arms, all receptor subtypes), while general control markers are receptor subtype-independent markers in the control arm. Results Global sensitivity biomarkers (n=42), where higher levels associate with pCR, included markers from mostly immune, HER2 and cell cycle/proliferation related pathways. In the TNBC population 13 sensitivity biomarkers were identified, which all overlapped with global sensitivity markers, e.g. immune gene signatures and EGFR protein. General control markers of sensitivity (n=15) included immune markers and cell cycle/proliferation signatures, amongst others, but were mainly distinct from sensitivity markers in the TNBC subgroup analysis of the control arm (n=6), with EGFR as the only overlapping biomarker. Remarkably, almost all general sensitivity markers in the control arm were also global sensitivity markers (n=13/15). Unlike sensitivity biomarkers, in TNBC only 2/11 biomarkers of resistance, where higher levels associate with non-pCR, overlapped with global resistance markers, which were cyclin D1 and p53. Global resistance markers (n=14) were mostly hormone receptor related biomarkers, while in TNBC resistance markers (n=11) included MEK1/2, FOXO1, CHK1 and CHK2, amongst others. Most significant general control markers of resistance were also global resistance markers (n=4/5), including IGF1R mRNA. Discussion This analysis shows that nearly all sensitivity and resistance biomarkers in the control arm overlap with global markers. We hypothesize that this is (partly) explained by the standard neoadjuvant chemotherapy backbone that is shared between all treatment arms. Interestingly, almost no resistance biomarkers in TNBC overlapped with global resistance markers, further supporting that resistance mechanisms in TNBC are receptor-subtype specific. Additional analysis of those markers could help further elucidate resistance biology in TNBC. Citation Format: Tam Binh V Bui, Denise M Wolf, Rosa I Gallagher, Julia Wulfkuhle, Christina Yau, Lamorna Brown-Swigart, Jennifer Rosenbluth, Douglas Yee, Gillian L Hirst, Laura J Esserman, Emanuel F Petricoin, Laura J Van 't Veer. Global and TNBC-specific biomarkers of sensitivity and resistance in early stage breast cancer from the I-SPY2 biomarker data compendium [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr A003.