The New Paradigm of Network Medicine to Analyze Breast Cancer Phenotypes

Prediction of hub genes and key pathways associated with the radiation response of human hematopoietic stem/progenitor cells using integrated bioinformatics methods

Hematopoietic stem cells (HSCs) are indispensable for the maintenance of the entire blood program through cytokine response. However, HSCs have high radiosensitivity, which is often a problem during radiation therapy and nuclear accidents. Although our previous study has reported that the combination cytokine treatment (interleukin-3, stem cell factor, and thrombopoietin) improves the survival of human hematopoietic stem/progenitor cells (HSPCs) after radiation, the mechanism by which cytokines contribute to the survival of HSPCs is largely unclear. To address this issue, the present study characterized the effect of cytokines on the radiation-induced gene expression profile of human CD34⁺ HSPCs and explored the hub genes that play key pathways associated with the radiation response using a cDNA microarray, a protein-protein interaction-MCODE module analysis and Cytohubba plugin tool in Cytoscape. This study identified 2,733 differentially expressed genes (DEGs) and five hub genes (TOP2A, EZH2, HSPA8, GART, HDAC1) in response to radiation in only the presence of cytokines. Furthermore, functional enrichment analysis found that hub genes and top DEGs based on fold change were enriched in the chromosome organization and organelle organization. The present findings may help predict the radiation response and improve our understanding of this response of human HSPCs.

Three genes expressed in relation to lipid metabolism considered as potential biomarkers for the diagnosis and treatment of diabetic peripheral neuropathy

Diabetic neuropathy is one of the most common chronic complications and is present in approximately 50% of diabetic patients. A bioinformatic approach was used to analyze candidate genes involved in diabetic distal symmetric polyneuropathy and their potential mechanisms. GSE95849 was downloaded from the Gene Expression Omnibus database for differential analysis, together with the identified diabetic peripheral neuropathy-associated genes and the three major metabolism-associated genes in the CTD database to obtain overlapping Differentially Expressed Genes (DEGs). Gene Set Enrichment Analysis and Functional Enrichment Analysis were performed. Protein-Protein Interaction and hub gene networks were constructed using the STRING database and Cytoscape software. The expression levels of target genes were evaluated using GSE24290 samples, followed by Receiver operating characteristic, curve analysis. And Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were performed on the target genes. Finally, mRNA-miRNA networks were constructed. A total of 442 co-expressed DEGs were obtained through differential analysis, of which 353 expressed up-regulated genes and 89 expressed down-regulated genes. The up-regulated DEGs were involved in 742 GOs and 10 KEGG enrichment results, mainly associated with lipid metabolism-related pathways, TGF-β receptor signaling pathway, lipid transport, and PPAR signaling pathway. A total of 4 target genes (CREBBP, EP300, ME1, CD36) were identified. Analysis of subject operating characteristic curves indicated that CREBBP (AUC = 1), EP300 (AUC = 0.917), ME1 (AUC = 0.944) and CD36 (AUC = 1) may be candidate serum biomarkers for DPN. Conclusion: Diabetic peripheral neuropathy pathogenesis and progression is caused by multiple pathways, which also provides clinicians with potential therapeutic tools.

Landscape of co-expressed genes between the myocardium and blood in sepsis and ceRNA network construction: a bioinformatic approach

Septic cardiomyopathy is a serious complication of sepsis. The mechanism of disease pathogenesis, which is caused by infection, is well researched. Despite ongoing efforts, there are no viable biological markers in the peripheral blood for early detection and diagnosis of septic cardiomyopathy. We aimed to uncover potential biomarkers of septic cardiomyopathy by comparing the covaried genes and pathways in the blood and myocardium of sepsis patients. Gene expression profiling of GSE79962, GSE65682, GSE54514, and GSE134364 was retrieved from the GEO database. Student's t-test was used for differential expression analysis. K-means clustering analysis was applied for subgroup identification. Least absolute shrinkage and selection operator (LASSO) and logistic regression were utilized for screening characteristic genes and model construction. Receiver operating characteristic (ROC) curves were generated for estimating the diagnostic efficacy. For ceRNA information prediction, miWalk and lncBase were applied. Cytoscape was used for ceRNA network construction. Inflammation-associated genes were upregulated, while genes related to mitochondria and aerobic metabolism were downregulated in both blood and the myocardium. Three groups with a significantly different mortality were identified by these covaried genes, using clustering analysis. Five characteristic genes-BCL2A1, CD44, ADGRG1, TGIF1, and ING3-were identified, which enabled the prediction of mortality of sepsis. The pathophysiological changes in the myocardium of patients with sepsis were also reflected in peripheral blood to some extent. The co-occurring pathological processes can affect the prognosis of sepsis. Thus, the genes we identified have the potential to become biomarkers for septic cardiomyopathy.

Identification of prognostic genes for early basal-like breast cancer with weighted gene co-expression network analysis

BACKGROUND

Breast cancer (BC) has become the leading cause of death for women's malignancies and increasingly threatens the health of women worldwide. However, there is a lack of effective targeted drugs for basal-like BC. Therefore, biomarkers related to the prognosis of early BC need to be identified.

METHODS

The RNA-seq data of 87 cases of early basal-like BC and 111 cases of normal breast tissue from The Cancer Genome Atlas were explored by the weighted gene co-expression network analysis method and Limma package. Then, intersected genes were identified, and hub genes were selected by the maximal clique centrality method. The prognostic effect of the hub genes was also evaluated in early basal-like BC.

RESULTS

In total, 601 IGs were identified in this study. An APPI network was constructed, and the top 10 hub genes were selected, namely, cyclin B1, cyclin A2, cyclin-dependent kinase 1, cell division cycle 20, DNA topoisomerase II alpha, BUB1 mitotic checkpoint serine/threonine kinase, aurora kinase B (AURKB), cyclin B2, kinesin family member 11, and assembly factor for spindle microtubules. Only AURKB was found to be significantly associated with the overall prognosis of early basal-like BC. The immune cell infiltration analysis showed that the infiltration numbers of CD4 + T cells and naïve CD8 + T cells were positively correlated with the AURKB expression level, while those of naïve B cells and macrophage M2 cells were negatively correlated with the AURKB expression level in basal-like BC.

CONCLUSION

AURKB might be a potential prognostic indicator in early basal-like BC.

A Review of the Nucleic Acid-Based Lateral Flow Assay for Detection of Breast Cancer from Circulating Biomarkers at a Point-of-Care in Low Income Countries

The current levels of breast cancer in African women have contributed to the high mortality rates among them. In South Africa, the incidence of breast cancer is also on the rise due to changes in behavioural and biological risk factors. Such low survival rates can be attributed to the late diagnosis of the disease due to a lack of access and the high costs of the current diagnostic tools. Breast cancer is asymptomatic at early stages, which is the best time to detect it and intervene to prevent high mortality rates. Proper risk assessment, campaigns, and access to adequate healthcare need to be prioritised among patients at an early stage. Early detection of breast cancer can significantly improve the survival rate of breast cancer patients, since therapeutic strategies are more effective at this stage. Early detection of breast cancer can be achieved by developing devices that are simple, sensitive, low-cost, and employed at point-of-care (POC), especially in low-income countries (LICs). Nucleic-acid-based lateral flow assays (NABLFAs) that combine molecular detection with the immunochemical visualisation principles, have recently emerged as tools for disease diagnosis, even for low biomarker concentrations. Detection of circulating genetic biomarkers in non-invasively collected biological fluids with NABLFAs presents an appealing and suitable method for POC testing in resource-limited regions and/or LICs. Diagnosis of breast cancer at an early stage will improve the survival rates of the patients. This review covers the analysis of the current state of NABLFA technologies used in developing countries to reduce the scourge of breast cancer.

A Comparison of Network-Based Methods for Drug Repurposing along with an Application to Human Complex Diseases

Drug repurposing strategy, proposing a therapeutic switching of already approved drugs with known medical indications to new therapeutic purposes, has been considered as an efficient approach to unveil novel drug candidates with new pharmacological activities, significantly reducing the cost and shortening the time of de novo drug discovery. Meaningful computational approaches for drug repurposing exploit the principles of the emerging field of Network Medicine, according to which human diseases can be interpreted as local perturbations of the human interactome network, where the molecular determinants of each disease (disease genes) are not randomly scattered, but co-localized in highly interconnected subnetworks (disease modules), whose perturbation is linked to the pathophenotype manifestation. By interpreting drug effects as local perturbations of the interactome, for a drug to be on-target effective against a specific disease or to cause off-target adverse effects, its targets should be in the nearby of disease-associated genes. Here, we used the network-based proximity measure to compute the distance between the drug module and the disease module in the human interactome by exploiting five different metrics (minimum, maximum, mean, median, mode), with the aim to compare different frameworks for highlighting putative repurposable drugs to treat complex human diseases, including malignant breast and prostate neoplasms, schizophrenia, and liver cirrhosis. Whilst the standard metric (that is the minimum) for the network-based proximity remained a valid tool for efficiently screening off-label drugs, we observed that the other implemented metrics specifically predicted further interesting drug candidates worthy of investigation for yielding a potentially significant clinical benefit.

Network analysis of atherosclerotic genes elucidates druggable targets

Background

Atherosclerosis is one of the major causes of cardiovascular disease. It is characterized by the accumulation of atherosclerotic plaque in arteries under the influence of inflammatory responses, proliferation of smooth muscle cell, accumulation of modified low density lipoprotein. The pathophysiology of atherosclerosis involves the interplay of a number of genes and metabolic pathways. In traditional translation method, only a limited number of genes and pathways can be studied at once. However, the new paradigm of network medicine can be explored to study the interaction of a large array of genes and their functional partners and their connections with the concerned disease pathogenesis. Thus, in our study we employed a branch of network medicine, gene network analysis as a tool to identify the most crucial genes and the miRNAs that regulate these genes at the post transcriptional level responsible for pathogenesis of atherosclerosis.

Result

From NCBI database 988 atherosclerotic genes were retrieved. The protein–protein interaction using STRING database resulted in 22,693 PPI interactions among 872 nodes (genes) at different confidence score. The cluster analysis of the 872 genes using MCODE, a plug-in of Cytoscape software revealed a total of 18 clusters, the topological parameter and gene ontology analysis facilitated in the selection of four influential genes viz., AGT, LPL, ITGB2, IRS1 from cluster 3. Further, the miRNAs (miR-26, miR-27, and miR-29 families) targeting these genes were obtained by employing MIENTURNET webtool.

Conclusion

Gene network analysis assisted in filtering out the 4 probable influential genes and 3 miRNA families in the pathogenesis of atherosclerosis. These genes, miRNAs can be targeted to restrict the occurrence of atherosclerosis. Given the importance of atherosclerosis, any approach in the understanding the genes involved in its pathogenesis can substantially enhance the health care system.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12920-022-01195-y.

In silico recognition of a prognostic signature in basal-like breast cancer patients

Background

Triple-negative breast cancers (TNBCs) display poor prognosis, have a high risk of tumour recurrence, and exhibit high resistance to drug treatments. Based on their gene expression profiles, the majority of TNBCs are classified as basal-like breast cancers. Currently, there are not available widely-accepted prognostic markers to predict outcomes in basal-like subtype, so the selection of new prognostic indicators for this BC phenotype represents an unmet clinical challenge.

Results

Here, we attempted to address this challenging issue by exploiting a bioinformatics pipeline able to integrate transcriptomic, genomic, epigenomic, and clinical data freely accessible from public repositories. This pipeline starts from the application of the well-established network-based SWIM methodology on the transcriptomic data to unveil important (switch) genes in relation with a complex disease of interest. Then, survival and linear regression analyses are performed to associate the gene expression profiles of the switch genes with both the patients’ clinical outcome and the disease aggressiveness. This allows us to identify a prognostic gene signature that in turn is fed to the last step of the pipeline consisting of an analysis at DNA level, to investigate whether variations in the expression of identified prognostic switch genes could be related to genetic (copy number variations) or epigenetic (DNA methylation differences) alterations in their gene loci, or to the activities of transcription factors binding to their promoter regions. Finally, changes in the protein expression levels corresponding to the so far identified prognostic switch genes are evaluated by immunohistochemical staining results taking advantage of the Human Protein Atlas.

Conclusion

The application of the proposed pipeline on the dataset of The Cancer Genome Atlas (TCGA)-Breast Invasive Carcinoma (BRCA) patients affected by basal-like subtype led to an in silico recognition of a basal-like specific gene signature composed of 11 potential prognostic biomarkers to be further investigated.

SWIMmeR: an R-based software to unveiling crucial nodes in complex biological networks

SUMMARY

We present SWIMmeR, an open-source version of its predecessor SWIM (SWitchMiner) that is a network-based tool for mining key (switch) genes that are associated with intriguing patterns of molecular co-abundance and may play a crucial role in phenotypic transitions in various biological settings. SWIM was originally written in MATLAB®, a proprietary programming language that requires the purchase of a license to install, manipulate, operate and run the software. Over the last years, SWIM has sparked a widespread interest within the scientific community thanks to the promising results obtained through its application in a broad range of phenotype-specific scenarios, spanning from complex diseases to grapevine berry maturation. This success has created the call for it to be distributed in a freely accessible, open-source, runtime environment, such as R, aimed at a general audience of non-expert users that cannot afford the leading proprietary solution. SWIMmeR is provided as a comprehensive collection of R functions and it also includes several additional features that make it less intensive in terms of computer time and more efficient in terms of usability and further implementation and extension.

AVAILABILITY AND IMPLEMENTATION

The SWIMmeR source code is freely available at https://github.com/sportingCode/SWIMmeR.git, along with a practical user guide, including a usage example of its application on breast cancer dataset.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Novel cancer subtyping method based on patient-specific gene regulatory network

The identification of cancer subtypes is important for the understanding of tumor heterogeneity. In recent years, numerous computational methods have been proposed for this problem based on the multi-omics data of patients. It is widely accepted that different cancer subtypes are induced by different molecular regulatory networks. However, only a few incorporate the differences between their molecular systems into the identification processes. In this study, we present a novel method to identify cancer subtypes based on patient-specific molecular systems. Our method realizes this by quantifying patient-specific gene networks, which are estimated from their transcriptome data, and by clustering their quantified networks. Comprehensive analyses of The Cancer Genome Atlas (TCGA) datasets applied to our method confirmed that they were able to identify more clinically meaningful cancer subtypes than the existing subtypes and found that the identified subtypes comprised different molecular features. Our findings also show that the proposed method can identify the novel cancer subtypes even with single omics data, which cannot otherwise be captured by existing methods using multi-omics data.

Improved supervised prediction of aging-related genes via weighted dynamic network analysis

BACKGROUND

This study focuses on the task of supervised prediction of aging-related genes from -omics data. Unlike gene expression methods for this task that capture aging-specific information but ignore interactions between genes (i.e., their protein products), or protein-protein interaction (PPI) network methods for this task that account for PPIs but the PPIs are context-unspecific, we recently integrated the two data types into an aging-specific PPI subnetwork, which yielded more accurate aging-related gene predictions. However, a dynamic aging-specific subnetwork did not improve prediction performance compared to a static aging-specific subnetwork, despite the aging process being dynamic. This could be because the dynamic subnetwork was inferred using a naive Induced subgraph approach. Instead, we recently inferred a dynamic aging-specific subnetwork using a methodologically more advanced notion of network propagation (NP), which improved upon Induced dynamic aging-specific subnetwork in a different task, that of unsupervised analyses of the aging process.

RESULTS

Here, we evaluate whether our existing NP-based dynamic subnetwork will improve upon the dynamic as well as static subnetwork constructed by the Induced approach in the considered task of supervised prediction of aging-related genes. The existing NP-based subnetwork is unweighted, i.e., it gives equal importance to each of the aging-specific PPIs. Because accounting for aging-specific edge weights might be important, we additionally propose a weighted NP-based dynamic aging-specific subnetwork. We demonstrate that a predictive machine learning model trained and tested on the weighted subnetwork yields higher accuracy when predicting aging-related genes than predictive models run on the existing unweighted dynamic or static subnetworks, regardless of whether the existing subnetworks were inferred using NP or the Induced approach.

CONCLUSIONS

Our proposed weighted dynamic aging-specific subnetwork and its corresponding predictive model could guide with higher confidence than the existing data and models the discovery of novel aging-related gene candidates for future wet lab validation.

Current and Future Therapies for Immunogenic Cell Death and Related Molecules to Potentially Cure Primary Breast Cancer

Simple Summary

How a cure for primary breast cancer after (neo)adjuvant therapy can be achieved at the molecular level remains unclear. Immune activation by anticancer drugs may contribute to the eradication of residual tumor cells by postoperative (neo)adjuvant chemotherapy. In addition, chemotherapy-induced immunogenic cell death (ICD) may result in long-term immune activation by memory effector T cells, leading to the curing of primary breast cancer. In this review, we discuss the molecular mechanisms by which anticancer drugs induce ICD and immunogenic modifications for antitumor immunity and targeted therapy against damage-associated molecular patterns. Our aim was to gain a better understanding of how to eradicate residual tumor cells treated with anticancer drugs and cure primary breast cancer by enhancing antitumor immunity with immune checkpoint inhibitors and vaccines.

Abstract

How primary breast cancer can be cured after (neo)adjuvant therapy remains unclear at the molecular level. Immune activation by anticancer agents may contribute to residual tumor cell eradication with postsurgical (neo)adjuvant chemotherapy. Chemotherapy-induced immunogenic cell death (ICD) may result in long-term immune activation with memory effector T cells, leading to a primary breast cancer cure. Anthracycline and taxane treatments cause ICD and immunogenic modulations, resulting in the activation of antitumor immunity through damage-associated molecular patterns (DAMPs), such as adenosine triphosphate, calreticulin, high mobility group box 1, heat shock proteins 70/90, and annexin A1. This response may eradicate residual tumor cells after surgical treatment. Although DAMP release is also implicated in tumor progression, metastasis, and drug resistance, thereby representing a double-edged sword, robust immune activation by anticancer agents and the subsequent acquisition of long-term antitumor immune memory can be essential components of the primary breast cancer cure. This review discusses the molecular mechanisms by which anticancer drugs induce ICD and immunogenic modifications for antitumor immunity and targeted anti-DAMP therapy. Our aim was to improve the understanding of how to eradicate residual tumor cells treated with anticancer drugs and cure primary breast cancer by enhancing antitumor immunity with immune checkpoint inhibitors and vaccines.

Identification of fibronectin 1 (FN1) and complement component 3 (C3) as immune infiltration-related biomarkers for diabetic nephropathy using integrated bioinformatic analysis

Immune cell infiltration (ICI) plays a pivotal role in the development of diabetic nephropathy (DN). Evidence suggests that immune-related genes play an important role in the initiation of inflammation and the recruitment of immune cells. However, the underlying mechanisms and immune-related biomarkers in DN have not been elucidated. Therefore, this study aimed to explore immune-related biomarkers in DN and the underlying mechanisms using bioinformatic approaches. In this study, four DN glomerular datasets were downloaded, merged, and divided into training and test cohorts. First, we identified 55 differentially expressed immune-related genes; their biological functions were mainly enriched in leukocyte chemotaxis and neutrophil migration. The CIBERSORT algorithm was then used to evaluate the infiltrated immune cells; macrophages M1/M2, T cells CD8, and resting mast cells were strongly associated with DN. The ICI-related gene modules as well as 25 candidate hub genes were identified to construct a protein-protein interactive network and conduct molecular complex detection using the GOSemSim algorithm. Consequently, FN1, C3, and VEGFC were identified as immune-related biomarkers in DN, and a related transcription factor-miRNA-target network was constructed. Receiver operating characteristic curve analysis was estimated in the test cohort; FN1 and C3 had large area under the curve values (0.837 and 0.824, respectively). Clinical validation showed that FN1 and C3 were negatively related to the glomerular filtration rate in patients with DN. Six potential therapeutic small molecule compounds, such as calyculin, phenamil, and clofazimine, were discovered in the connectivity map. In conclusion, FN1 and C3 are immune-related biomarkers of DN.

In silico drug repurposing in COVID-19: A network-based analysis

The SARS-CoV-2 pandemic is a worldwide public health emergency. Despite the beginning of a vaccination campaign, the search for new drugs to appropriately treat COVID-19 patients remains a priority. Drug repurposing represents a faster and cheaper method than de novo drug discovery. In this study, we examined three different network-based approaches to identify potentially repurposable drugs to treat COVID-19. We analyzed transcriptomic data from whole blood cells of patients with COVID-19 and 21 other related conditions, as compared with those of healthy subjects. In addition to conventionally used drugs (e.g., anticoagulants, antihistaminics, anti-TNFα antibodies, corticosteroids), unconventional candidate compounds, such as SCN5A inhibitors and drugs active in the central nervous system, were identified. Clinical judgment and validation through clinical trials are always mandatory before use of the identified drugs in a clinical setting.

Expression levels of chemokine (C-X-C motif) ligands CXCL1 and CXCL3 as prognostic biomarkers in rectal adenocarcinoma: evidence from Gene Expression Omnibus (GEO) analyses

Rectal cancer is a life‑threatening disease worldwide. Chemotherapy resistance is common in rectal adenocarcinoma patients and has unfavorable survival outcomes; however, its related molecular mechanisms remain unknown. To identify genes related to the initiation and progression of rectal adenocarcinoma, three datasets were obtained from the Gene Expression Omnibus database. In total, differentially expressed genes were analyzed from 294 tumor and 277 para-carcinoma samples from patients with rectal cancer. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes functions were investigated. Cytoscape software and MicroRNA Enrichment Turned Network were applied to construct a protein-protein interaction network of the dependent hub genes and related microRNAs. The Oncomine database was used to identify hub genes. Additionally, Gene Expression Profiling Interactive Analysis was applied to determine the RNA expression level. Tumor immune infiltration was assessed using the Tumor Immune Estimation Resource database. The expression profiles of hub genes between stages, and their prognostic value, were also evaluated. During this study, data from The Cancer Genome Atlas were utilized. In rectal adenocarcinoma, four hub genes including CXCL1, CXCL2, CXCL3, and GNG4 were highly expressed at the gene and RNA levels. The expression of CXCL1, CXCL2, and CXCL3 was regulated by has-miR-1-3p and had a strong positive correlation with macrophage and neutrophil. CXCL2 and CXCL3 were differentially expressed at different tumor stages. High expression levels of CXCL1 and CXCL3 predicted poor survival. In conclusion, the CXCL1 and CXCL3 genes may have potential for prognosis and molecular targeted therapy of rectal adenocarcinoma.

The risk of pancreatic adenocarcinoma following SARS-CoV family infection

COVID 19 disease has become a global catastrophe over the past year that has claimed the lives of over two million people around the world. Despite the introduction of vaccines against the disease, there is still a long way to completely eradicate it. There are concerns about the complications following infection with SARS-CoV-2. This research aimed to evaluate the possible correlation between infection with SARS-CoV viruses and cancer in an in-silico study model. To do this, the relevent dataset was selected from GEO database. Identification of differentially expressed genes among defined groups including SARS-CoV, SARS-dORF6, SARS-BatSRBD, and H1N1 were screened where the |Log FC| ≥ 1and p < 0.05 were considered statistically significant. Later, the pathway enrichment analysis and gene ontology (GO) were used by Enrichr and Shiny GO databases. Evaluation with STRING online was applied to predict the functional interactions of proteins, followed by Cytoscape analysis to identify the master genes. Finally, analysis with GEPIA2 server was carried out to reveal the possible correlation between candidate genes and cancer development. The results showed that the main molecular function of up- and down-regulated genes was "double-stranded RNA binding" and actin-binding, respectively. STRING and Cytoscape analysis presented four genes, PTEN, CREB1, CASP3, and SMAD3 as the key genes involved in cancer development. According to TCGA database results, these four genes were up-regulated notably in pancreatic adenocarcinoma. Our findings suggest that pancreatic adenocarcinoma is the most probably malignancy happening after infection with SARS-CoV family.

Gene network analysis using SWIM reveals interplay between the transcription factor-encoding genes HMGA1, FOXM1, and MYBL2 in triple-negative breast cancer

Among breast cancer subtypes, triple-negative breast cancer (TNBC) is the most aggressive with the worst prognosis and the highest rates of metastatic disease. To identify TNBC gene signatures, we applied the network-based methodology implemented by the SWIM software to gene expression data of TNBC patients in The Cancer Genome Atlas (TCGA) database. SWIM enables to predict key (switch) genes within the co-expression network, whose perturbations in expression pattern and abundance may contribute to the (patho)biological phenotype. Here, SWIM analysis revealed an interesting interplay between the genes encoding the transcription factors HMGA1, FOXM1, and MYBL2, suggesting a potential cooperation among these three switch genes in TNBC development. The correlative nature of this interplay in TNBC was assessed by in vitro experiments, demonstrating how they may actually modulate the expression of each other.

MuSA: a graphical user interface for multi-OMICs data integration in radiogenomic studies

Analysis of large-scale omics data along with biomedical images has gaining a huge interest in predicting phenotypic conditions towards personalized medicine. Multiple layers of investigations such as genomics, transcriptomics and proteomics, have led to high dimensionality and heterogeneity of data. Multi-omics data integration can provide meaningful contribution to early diagnosis and an accurate estimate of prognosis and treatment in cancer. Some multi-layer data structures have been developed to integrate multi-omics biological information, but none of these has been developed and evaluated to include radiomic data. We proposed to use MultiAssayExperiment (MAE) as an integrated data structure to combine multi-omics data facilitating the exploration of heterogeneous data. We improved the usability of the MAE, developing a Multi-omics Statistical Approaches (MuSA) tool that uses a Shiny graphical user interface, able to simplify the management and the analysis of radiogenomic datasets. The capabilities of MuSA were shown using public breast cancer datasets from TCGA-TCIA databases. MuSA architecture is modular and can be divided in Pre-processing and Downstream analysis. The pre-processing section allows data filtering and normalization. The downstream analysis section contains modules for data science such as correlation, clustering (i.e., heatmap) and feature selection methods. The results are dynamically shown in MuSA. MuSA tool provides an easy-to-use way to create, manage and analyze radiogenomic data. The application is specifically designed to guide no-programmer researchers through different computational steps. Integration analysis is implemented in a modular structure, making MuSA an easily expansible open-source software.

An Integrative Computational Approach Based on Expression Similarity Signatures to Identify Protein-Protein Interaction Networks in Female-Specific Cancers

Breast, ovarian, and endometrial cancers have a major impact on mortality in women. These tumors share hormone-dependent mechanisms involved in female-specific cancers which support tumor growth in a different manner. Integrated computational approaches may allow us to better detect genomic similarities between these different female-specific cancers, helping us to deliver more sophisticated diagnosis and precise treatments. Recently, several initiatives of The Cancer Genome Atlas (TCGA) have encouraged integrated analyses of multiple cancers rather than individual tumors. These studies revealed common genetic alterations (driver genes) even in clinically distinct entities such as breast, ovarian, and endometrial cancers. In this study, we aimed to identify expression similarity signatures by extracting common genes among TCGA breast (BRCA), ovarian (OV), and uterine corpus endometrial carcinoma (UCEC) cohorts and infer co-regulatory protein-protein interaction networks that might have a relationship with the estrogen signaling pathway. Thus, we carried out an unsupervised principal component analysis (PCA)-based computational approach, using RNA sequencing data of 2,015 female cancer and 148 normal samples, in order to simultaneously capture the data heterogeneity of intertumors. Firstly, we identified tumor-associated genes from gene expression profiles. Secondly, we investigated the signaling pathways and co-regulatory protein-protein interaction networks underlying these three cancers by leveraging the Ingenuity Pathway Analysis software. In detail, we discovered 1,643 expression similarity signatures (638 downregulated and 1,005 upregulated genes, with respect to normal phenotype), denoted as tumor-associated genes. Through functional genomic analyses, we assessed that these genes were involved in the regulation of cell-cycle-dependent mechanisms, including metaphase kinetochore formation and estrogen-dependent S-phase entry. Furthermore, we generated putative co-regulatory protein-protein interaction networks, based on upstream regulators such as the ERBB2/HER2 gene. Moreover, we provided an ad-hoc bioinformatics workflow with a manageable list of intertumor expression similarity signatures for the three female-specific cancers. The expression similarity signatures identified in this study might uncover potential estrogen-dependent molecular mechanisms promoting carcinogenesis.