Search

Background: High-grade serious ovarian carcinoma (HGSOC) is a subtype of ovarian cancer with a different prognosis attributable to genetic heterogeneity. The prognosis of patients with advanced HGSOC requires prediction by genetic markers. This study systematically analyzed gene expression profile data to establish a genetic marker for predicting HGSOC prognosis. Methods: The RNA-seq data set and information on clinical follow-up of HGSOC were retrieved from Gene Expression Omnibus (GEO) database, and the data were standardized by DESeq2 as a training set. On the other hand, HGSOC RNA sequence data and information on clinical follow-up were retrieved from The Cancer Genome Atlas (TCGA) as a test set. Additionally, ovarian cancer microarray data set was obtained from GEO as the external validation set. Prognostic genes were screened from the training set, and characteristic selection was performed using the least absolute shrinkage and selection operator (LASSO) with 80% re-sampling for 5000 times. Genes with a frequency of more than 2000 were selected as robust biomarkers. Finally, a gene-related prognostic model was validated in both the test and GEO validation sets. Results: A total of 148 genes were found to be significantly correlated with HGSOC prognosis. The expression profile of these genes could stratify HGSOC prognosis and they were enriched to multiple tumor-related regulatory pathways such as tyrosine metabolism and AMPK signaling pathway. AKR1B10 and ANGPT4 were obtained after 5000-time re-sampling by LASSO regression. AKR1B10 was associated with the metastasis and progression of several tumors. In this study, Cox regression analysis was performed to create a 2-gene signature as an independent prognostic factor for HGSOC, which has the ability to stratify risk samples in all three data sets (p < 0.05). The Gene Set Enrichment Analysis (GSEA) discovered abnormally active REGULATION_OF_AUTOPHAGY and OLFACTORY_TRANSDUCTION pathways in the high-risk group samples. Conclusion: This study resulted in the creation of a 2-gene molecular prognostic classifier that distinguished clinical features and was a promising novel prognostic tool for assessing the prognosis of HGSOC. RiskScore was a novel prognostic model which might be effective in guiding accurate prognosis of HGSOC. [ABSTRACT FROM AUTHOR]

Background: SH2 domain containing 1A (SH2D1A) expression has been linked to cancer progression. However, the functions of SH2D1A in hepatocellular carcinoma (HCC) have not been reported. Methods: The effects of SH2D1A on the proliferation, migration, and invasion of HCC cells and the related pathways were re-explored in cell models with SH2D1A overexpression using the CCK-8, migration and invasion assays and western blotting. The functions and mechanisms of genes co-expressed with SH2D1A were analyzed using gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. The relationship between SH2D1A expression and immune microenvironment features in HCC was explored. Results: Elevated SH2D1A expression promoted cell proliferation, migration, and invasion, which was related to the overexpression of p-Nf-κB and BCL2A1 protein levels in HCC. SH2D1A expression was related to the immune, stromal, and ESTIMATE scores, and the abundance of immune cells, such as B cells, CD8+ T cells, and T cells. SH2D1A expression was significantly related to the expression of immune cell markers, such as PDCD1, CD8A, and CTLA4 in HCC. Conclusion: SH2D1A overexpression was found to promote cell growth and metastasis via the Nf-κB signaling pathway and may be related to the immune microenvironment in HCC. The findings indicate that SH2D1A can function as a biomarker in HCC. [ABSTRACT FROM AUTHOR]

Background: Lung cancer is reported to be the leading cause of death in males and females, globally. Increasing evidence highlights the paramount importance of Lactate dehydrogenase D (LDHD) in different types of cancers, though it's role in lung adenocarcinoma (LUAD) is still inadequately explored. In this study, we aimed to investigate and determine the relationship between LDHD and LUAD. Methods: The collection of the samples was guided by The Cancer Genome Atlas (TCGA) datasets and Gene Expression Omnibus (GEO). To ascertain various aspects around LDHD function, we analyzed different expression genes (DEGs), functional enrichment, and protein–protein interaction (PPI) networks. The predictive values for LDHD were collectively determined using the Kaplan–Meier method, Cox regression analysis, and a nomogram. Evaluation of the immune infiltration analysis was completed using Estimate and ssGSEA. The prediction of the immunotherapy response was based on TIDE and IPS. The LDHD expression levels in LUAD were validated through Western blot, qPCR, and immunohistochemistry methods. Wound healing and transwell assays were also performed to illustrate the aggressive features in LUAD cell lines. Results: The results showed that LDHD was generally downregulated in LUAD patients, with the low LDHD group presenting a decline in OS, DSS, and PFI. Enriched pathways, which include pyruvate metabolism, central carbon metabolism, and oxidative phosphorylation were observed through KEGG analysis. It was also noted that the expression of LDHD expression was inversely related to immune cell infiltration and typical checkpoints. The high LDHD group's response to immunotherapy was remarkable, particularly in CTAL4 + /PD1- therapy. In vitro studies revealed that the overexpression of LDHD caused tumor migration and invasion to be suppressed. Conclusion: In conclusion, our study revealed that LDHD might be an effective predictor of prognosis and immune filtration, possibly leading to better choices for immunotherapy. [ABSTRACT FROM AUTHOR]

Background: The tumor immune microenvironment plays a crucial role in the efficacy of various therapeutics. However, their correlation is not yet completely understood in Clear cell renal cell carcinoma (ccRCC). This study aimed to investigate the potential of TREM-1 as a potential novel biomarker for ccRCC. Methods: We constructed a ccRCC immune prognostic signature. The clinical characteristics, the status of the tumor microenvironment, and immune infiltration were analyzed through the ESTIMATE and CIBERSORT algorithms for the hub gene, while the Gene Set Enrichment Analysis and PPI analysis were performed to predict the function of the hub gene. Immunohistochemical staining was used to detect the expression of TREM-1 in renal clear cell carcinoma tissues. Results: The CIBERSORT and ESTIMATE algorithms revealed that TREM-1 was correlated with the infiltration of 12 types of immune cells. Therefore, it was determined that TREM-1 was involved in numerous classical pathways in the immune response via GSEA analysis. In Immunohistochemical staining, we found that the expression of TREM-1 was significantly upregulated with increasing tumor grade in renal clear cell carcinoma, and elevated TREM-1 expression was associated with poor prognosis. Conclusions: The results suggest that TREM-1 may act as an implicit novel prognostic biomarker in ccRCC that could be utilized to facilitate immunotherapeutic strategy. [ABSTRACT FROM AUTHOR]

Background: Endometrial cancer (EC) with metastasis in pelvic/para-aortic lymph nodes suggests an unsatisfactory prognosis. Nevertheless, there is still rare literature focusing on the role of epithelial-mesenchymal transition (EMT) in lymph node metastasis (LNM) in EC. Methods: Transcriptional data were derived from the TCGA database. Patients with stage IA–IIIC2 EC were included, constituting the LN-positive and LN-negative groups. To evaluate the extent of EMT, an EMT signature composed of 315 genes was adopted. The EMT-related genes (ERGs) were obtained from the dbEMT2 database, and the differentially expressed ERGs (DEERGs) between these two groups were screened. On the basis of DEERGs, pathway analysis was carried out. We eventually adopted the logistic regression model to build an ERG-based gene signature with predictive value for LNM in EC. Results: A total of 498 patients were included, with 75 in the LN-positive group. Median EMT score of tumor tissues from LN-negative group was − 0.369, while that from the LN-positive group was − 0.296 (P < 0.001), which clearly exhibited a more mesenchymal phenotype for LNM cases on the EMT continuum. By comparing expression profiles, 266 genes were identified as DEERGs, in which 184 were upregulated and 82 were downregulated. In pathway analysis, various EMT-related pathways were enriched. DEERGs shared between molecular subtypes were comparatively few. The ROC curve and logistic regression analysis screened 7 genes with the best performance to distinguish between the LN-positive and LN-negative group, i.e., CIRBP, DDR1, F2RL2, HOXA10, PPARGC1A, SEMA3E, and TGFB1. A logistic regression model including the 7-gene-based risk score, age, grade, myometrial invasion, and histological subtype was built, with an AUC of 0.850 and a favorite calibration (P = 0.074). In the validation dataset composed of 83 EC patients, the model exhibited a satisfactory predictive value and was well-calibrated (P = 0.42). Conclusion: The EMT status and expression of ERGs varied in LNM and non-LNM EC tissues, involving multiple EMT-related signaling pathways. Aside from that, the distribution of DEERGs differed among molecular subtypes. An ERG-based gene signature including 7 DEERGs exhibited a desirable predictive value for LNM in EC, which required further validation based upon clinical specimens in the future. [ABSTRACT FROM AUTHOR]

Background: CXC chemokine receptor gene family consists of seven well-established members which are broadly involved in biological functions of various cancers. Currently, limited studies have shed light on the expression profile of CXCR family members (CXCRs), as well as their prognostic value, in head and neck squamous cells carcinoma (HNSCC). Methods: The data for this study were retrieved from the Cancer Genome Atlas database and other publicly available databases, including gene expression, methylation profiles, clinical information, immunological features, and prognoses. The expression pattern and prognostic values of CXCRs were identified, and the potential mechanism underlying CXCRs function in HNSCC was investigated by gene set enrichment analysis (GSEA). Results: CXCRs were differentially expressed in HNSCC. As shown by Kaplan–Meier analysis, high CXCR3-6 expression was significantly associated with better prognostic outcomes of HNSCC patients, including overall survival and progression-free survival. According to the results of univariate and multivariate Cox proportional risk regression analysis, it was demonstrated that upregulation of CXCR3-6 was an independent factor for better prognosis, while the two other clinical features, age and stage, were factors for worse prognosis. A significant positive correlation between CXCR3-6 and tumor-infiltrated immune cells was revealed by results from Tumor Immune Estimation Resource and CIBERSORT analysis database. The main involvement of CXCRs in immune and inflammatory responses was further confirmed by GSEA. Conclusions: Overall, this study provided a rationale for targeting CXCRs as a promising therapeutic strategy of HNSCC. [ABSTRACT FROM AUTHOR]

Background: Acute myeloid leukemia (AML) is a group of highly heterogenous and aggressive blood cancer. Despite recent progress in its diagnosis and treatment, patient outcome is variable and drug resistance results in increased mortality. The siglec family plays an important role in tumorigenesis and aging. Increasing age is a risk factor for AML and cellular aging contributes to leukemogenesis via various pathways. Methods: The differential expression of the siglec family was compared between 151 AML patients and 70 healthy controls, with their information downloaded from TCGA and GTEx databases, respectively. How siglec expression correlated to AML patient clinical features, immune cell infiltration, drug resistance and survival outcome was analyzed. Differentially expressed genes in AML patients with low- and high-expressed siglec9 and siglec14 were analyzed and functionally enriched. The aging-related gene set was merged with the differentially expressed genes in AML patients with low and high expression of siglec9, and merged genes were subjected to lasso regression analysis to construct a novel siglec-based and aging-related prognostic model. The prediction model was validated using a validation cohort from GEO database (GSE106291). Results: The expression levels of all siglec members were significantly altered in AML. The expression of siglecs was significantly correlated with AML patient clinical features, immune cell infiltration, drug resistance, and survival outcome. Based on the differentially expressed genes and aging-related gene set, we developed a 9-gene prognostic model and decision curve analysis revealed the net benefit generated by our prediction model. The siglec-based and aging-related 9-gene prognostic model was tested using a validation data set, in which AML patients with higher risk scores had significantly reduced survival probability. Time-dependent receiver operating characteristic curve and nomogram were plotted and showed the diagnostic accuracy and predictive value of our 9-gene prognostic model, respectively. Conclusions: Overall, our study indicates the important role of siglec family in AML and the good performance of our novel siglec-based and aging-related 9-gene signature in predicting AML patient outcome. [ABSTRACT FROM AUTHOR]

Background: Numerous studies have implicated autophagy in the pathogenesis of thyroid carcinoma. This investigation aimed to establish an autophagy-related gene model and nomogram that can help predict the overall survival (OS) of patients with differentiated thyroid carcinoma (DTHCA). Methods: Clinical characteristics and RNA-seq expression data from TCGA (The Cancer Genome Atlas) were used in the study. We also downloaded autophagy-related genes (ARGs) from the Gene Set Enrichment Analysis website and the Human Autophagy Database. First, we assigned patients into training and testing groups. R software was applied to identify differentially expressed ARGs for further construction of a protein-protein interaction (PPI) network for gene functional analyses. A risk score-based prognostic risk model was subsequently developed using univariate Cox regression and LASSO-penalized Cox regression analyses. The model's performance was verified using Kaplan-Meier (KM) survival analysis and ROC curve. Finally, a nomogram was constructed for clinical application in evaluating the patients with DTHCA. Finally, a 7-gene prognostic risk model was developed based on gene set enrichment analysis. Results: Overall, we identified 54 differentially expressed ARGs in patients with DTHCA. A new gene risk model based on 7-ARGs (CDKN2A, FGF7, CTSB, HAP1, DAPK2, DNAJB1, and ITPR1) was developed in the training group and validated in the testing group. The predictive accuracy of the model was reflected by the area under the ROC curve (AUC) values. Univariate and multivariate Cox regression analysis indicated that the model could independently predict the prognosis of patients with THCA. The constrained nomogram derived from the risk score and age also showed high prediction accuracy. Conclusions: Here, we developed a 7-ARG prognostic risk model and nomogram for differentiated thyroid carcinoma patients that can guide clinical decisions and individualized therapy. [ABSTRACT FROM AUTHOR]

Objectives: A novel graph data model of non-small cell lung cancer clinical and genomic data has been constructed with two aims: (1) provide a suitable model for facilitating graph analytics within the Neo4j framework or through tools which can interact through existing Neo4j APIs; and (2) provide a base model extensible to other cancer types and additional datasets such as those derived from electronic health records and other real world sources. Data description: Clinical and genomic data integrated with a novel property graph database schema from publicly available datasets and analyses based on The Cancer Genome Atlas lung cancer datasets augmented by with subgraphs patient-patient social network from similarity and correlation as well as individual based biological networks. [ABSTRACT FROM AUTHOR]

Background: Accumulating evidences demonstrated tumor microenvironment (TME) of bladder cancer (BLCA) may play a pivotal role in modulating tumorigenesis, progression, and alteration of biological features. Currently we aimed to establish a prognostic model based on TME-related gene expression for guiding clinical management of BLCA.Methods: We employed ESTIMATE algorithm to evaluate TME cell infiltration in BLCA. The RNA-Seq data from The Cancer Genome Atlas (TCGA) database was used to screen out differentially expressed genes (DEGs). Underlying relationship between co-expression modules and TME was investigated via Weighted gene co-expression network analysis (WGCNA). COX regression and the least absolute shrinkage and selection operator (LASSO) analysis were applied for screening prognostic hub gene and establishing a risk predictive model. BLCA specimens and adjacent tissues from patients were obtained from patients. Bladder cancer (T24, EJ-m3) and bladder uroepithelial cell line (SVHUC1) were used for genes validation. qRT-PCR was employed to validate genes mRNA level in tissues and cell lines.Results: 365 BLCA samples and 19 adjacent normal samples were selected for identifying DEGs. 2141 DEGs were identified and used to construct co-expression network. Four modules (magenta, brown, yellow, purple) were regarded as TME regulatory modules through WGCNA and GO analysis. Furthermore, seven hub genes (ACAP1, ADAMTS9, TAP1, IFIT3, FBN1, FSTL1, COL6A2) were screened out to establish a risk predictive model via COX and LASSO regression. Survival analysis and ROC curve analysis indicated our predictive model had good performance on evaluating patients prognosis in different subgroup of BLCA. qRT-PCR result showed upregulation of ACAP1, IFIT3, TAP1 and downregulation of ADAMTS9, COL6A2, FSTL1,FBN1 in BLCA specimens and cell lines.Conclusions: Our study firstly integrated multiple TME-related genes to set up a risk predictive model. This model could accurately predict BLCA progression and prognosis, which offers clinical implication for risk stratification, immunotherapy drug screen and therapeutic decision. [ABSTRACT FROM AUTHOR]

Background: The objectives of this study were to screen out cut-off age value and age-related differentially expressed genes (DEGs) in clear cell renal cell carcinoma (CCRCC) from Surveillance Epidemiology and End Results (SEER) database and The Cancer Genome Atlas (TCGA) database.Methods: We selected 45,974 CCRCC patients from SEER and 530 RNA-seq data from TCGA database. The age cut-off value was defined using the X-tile program. Propensity score matching (PSM) was used to balance the differences between young and old groups. Hazard ratio (HR) was applied to evaluate prognostic risk of age in different subgroups. Age-related DEGs were identified via RNA-seq data. Survival analysis was used to assess the relationship between DEGs and prognosis.Results: In this study, we divided the patients into young (n = 14,276) and old (n = 31,698) subgroups according to cut-off value (age = 53). Age > 53 years was indicated as independent risk factor for overall survival (OS) and cancer specific survival (CSS) of CCRCC before and after PSM. The prognosis of old group was worse than that in young group. Eleven gene were differential expression between the younger and older groups in CCRCC. The expression levels of PLA2G2A and SIX2 were related to prognosis of the elderly.Conclusion: Fifty-three years old was cut-off value in CCRCC. The prognosis of the elderly was worse than young people. It remind clinicians that more attention and better treatment should be given to CCRCC patients who are over 53 years old. PLA2G2A and SIX2 were age-related differential genes which might play an important role in the poor prognosis of elderly CCRCC patients. [ABSTRACT FROM AUTHOR]

Background: Recently, the possibility of tumour classification based on genetic data has been investigated. However, genetic datasets are difficult to handle because of their massive size and complexity of manipulation. In the present study, we examined the diagnostic performance of machine learning applications using imaging-based classifications of oral squamous cell carcinoma (OSCC) gene sets. Methods: RNA sequencing data from SCC tissues from various sites, including oral, non-oral head and neck, oesophageal, and cervical regions, were downloaded from The Cancer Genome Atlas (TCGA). The feature genes were extracted through a convolutional neural network (CNN) and machine learning, and the performance of each analysis was compared. Results: The ability of the machine learning analysis to classify OSCC tumours was excellent. However, the tool exhibited poorer performance in discriminating histopathologically dissimilar cancers derived from the same type of tissue than in differentiating cancers of the same histopathologic type with different tissue origins, revealing that the differential gene expression pattern is a more important factor than the histopathologic features for differentiating cancer types. Conclusion: The CNN-based diagnostic model and the visualisation methods using RNA sequencing data were useful for correctly categorising OSCC. The analysis showed differentially expressed genes in multiwise comparisons of various types of SCCs, such as KCNA10, FOSL2, and PRDM16, and extracted leader genes from pairwise comparisons were FGF20, DLC1, and ZNF705D. [ABSTRACT FROM AUTHOR]

Background: Few studies have examined prognostic outcomes-associated molecular signatures other than overall survival (OS) for gastric cancer (GC). We aimed to identify DNA methylation biomarkers associated with multiple prognostic outcomes of GC in an epigenome-wide association study.Methods: Based on the Cancer Genome Atlas (TCGA), DNA methylation loci associated with OS (n = 381), disease-specific survival (DSS, n = 372), and progression-free interval (PFI, n = 383) were discovered in training set subjects (false discovery rates < 0.05) randomly selected for each prognostic outcome and were then validated in remaining subjects (P-values < 0.05). Key CpGs simultaneously validated for OS, DSS, and PFI were further assessed for disease-free interval (DFI, n = 247). Gene set enrichment analyses were conducted to explore the Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathways simultaneously enriched for multiple GC prognostic outcomes. Methylation correlated blocks (MCBs) were identified for co-methylation patterns associated with GC prognosis. Based on key CpGs, risk score models were established to predict four prognostic outcomes. Spearman correlation analyses were performed between key CpG sites and their host gene mRNA expression.Results: We newly identified DNA methylation of seven CpGs significantly associated with OS, DSS, and PFI of GC, including cg10399824 (GRK5), cg05275153 (RGS12), cg24406668 (MMP9), cg14719951(DSC3), and cg25117092 (MED12L), and two in intergenic regions (cg11348188 and cg11671115). Except cg10399824 and cg24406668, five of them were also significantly associated with DFI of GC. Neuroactive ligand-receptor interaction pathway was suggested to play a key role in the effect of DNA methylation on GC prognosis. Consistent with individual CpG-level association, three MCBs involving cg11671115, cg14719951, and cg24406668 were significantly associated with multiple prognostic outcomes of GC. Integrating key CpG loci, two risk score models performed well in predicting GC prognosis. Gene body DNA methylation of cg14719951, cg10399824, and cg25117092 was associated with their host gene expression, whereas no significant associations between their host gene expression and four clinical prognostic outcomes of GC were observed.Conclusions: We newly identified seven CpGs associated with OS, DSS, and PFI of GC, with five of them also associated with DFI, which might inform patient stratification in clinical practices. [ABSTRACT FROM AUTHOR]

Background: Breast cancer (BRCA) is one of the most common cancers worldwide. Abnormal alternative splicing (AS) frequently observed in cancers. This study aims to demonstrate AS events and signatures that might serve as prognostic indicators for BRCA.Methods: Original data for all seven types of splice events were obtained from TCGA SpliceSeq database. RNA-seq and clinical data of BRCA cohorts were downloaded from TCGA database. Survival-associated AS events in BRCA were analyzed by univariate COX proportional hazards regression model. Prognostic signatures were constructed for prognosis prediction in patients with BRCA based on survival-associated AS events. Pearson correlation analysis was performed to measure the correlation between the expression of splicing factors (SFs) and the percent spliced in (PSI) values of AS events. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were conducted to demonstrate pathways in which survival-associated AS event is enriched.Results: A total of 45,421 AS events in 21,232 genes were identified. Among them, 1121 AS events in 931 genes significantly correlated with survival for BRCA. The established AS prognostic signatures of seven types could accurately predict BRCA prognosis. The comprehensive AS signature could serve as independent prognostic factor for BRCA. A SF-AS regulatory network was therefore established based on the correlation between the expression levels of SFs and PSI values of AS events.Conclusions: This study revealed survival-associated AS events and signatures that may help predict the survival outcomes of patients with BRCA. Additionally, the constructed SF-AS networks in BRCA can reveal the underlying regulatory mechanisms in BRCA. [ABSTRACT FROM AUTHOR]

Background: This study aimed to determine and verify the prognostic value and potential functional mechanism of signal recognition particle 14 (SRP14) in acute myeloid leukemia (AML) using a genome-wide expression profile dataset. Methods: We obtained an AML genome-wide expression profile dataset and clinical prognostic data from The Cancer Genome Atlas (TCGA) and GSE12417 databases, and explored the prognostic value and functional mechanism of SRP14 in AML using survival analysis and various online tools. Results: Survival analysis showed that AML patients with high SRP14 expression had poorer overall survival than patients with low SRP14 expression. Time-dependent receiver operating characteristic curves indicated that SRP14 had good accuracy for predicting the prognosis in patients with AML. Genome-wide co-expression analysis suggested that SRP14 may play a role in AML by participating in the regulation of biological processes and signaling pathways, such as cell cycle, cell adhesion, mitogen-activated protein kinase, tumor necrosis factor, T cell receptor, DNA damage response, and nuclear factor-kappa B (NF-κB) signaling. Gene set enrichment analysis indicated that SRP14 was significantly enriched in biological processes and signaling pathways including regulation of hematopoietic progenitor cell differentiation and stem cell differentiation, intrinsic apoptotic signaling pathway by p53 class mediator, interleukin-1, T cell mediated cytotoxicity, and NF-κB-inducing kinase/NF-κB signaling. Using the TCGA AML dataset, we also identified four drugs (phenazone, benzydamine, cinnarizine, antazoline) that may serve as SRP14-targeted drugs in AML. Conclusion: The current results revealed that high SRP14 expression was significantly related to a poor prognosis and may serve as a prognostic biomarker in patients with AML. [ABSTRACT FROM AUTHOR]

Background: Given that dysregulated metabolism has been recently identified as a hallmark of cancer biology, this study aims to establish and validate a prognostic signature of lung adenocarcinoma (LUAD) based on metabolism-related genes (MRGs). Methods: The gene sequencing data of LUAD samples with clinical information and the metabolism-related gene set were obtained from The Cancer Genome Atlas (TCGA) and Molecular Signatures Database (MSigDB), respectively. The differentially expressed MRGs were identified by Wilcoxon rank sum test. Then, univariate cox regression analysis was performed to identify MRGs that related to overall survival (OS). A prognostic signature was developed by multivariate Cox regression analysis. Furthermore, the signature was validated in the GSE31210 dataset. In addition, a nomogram that combined the prognostic signature was created for predicting the 1-, 3- and 5-year OS of LUAD. The accuracy of the nomogram prediction was evaluated using a calibration plot. Finally, cox regression analysis was applied to identify the prognostic value and clinical relationship of the signature in LUAD. Results: A total of 116 differentially expressed MRGs were detected in the TCGA dataset. We found that 12 MRGs were most significantly associated with OS by using the univariate regression analysis in LUAD. Then, multivariate Cox regression analyses were applied to construct the prognostic signature, which consisted of six MRGs-aldolase A (ALDOA), catalase (CAT), ectonucleoside triphosphate diphosphohydrolase-2 (ENTPD2), glucosamine-phosphate N-acetyltransferase 1 (GNPNAT1), lactate dehydrogenase A (LDHA), and thymidylate synthetase (TYMS). The prognostic value of this signature was further successfully validated in the GSE31210 dataset. Furthermore, the calibration curve of the prognostic nomogram demonstrated good agreement between the predicted and observed survival rates for each of OS. Further analysis indicated that this signature could be an independent prognostic indicator after adjusting to other clinical factors. The high-risk group patients have higher levels of immune checkpoint molecules and are therefore more sensitive to immunotherapy. Finally, we confirmed six MRGs protein and mRNA expression in six lung cancer cell lines and firstly found that ENTPD2 might played an important role on LUAD cells colon formation and migration. Conclusions: We established a prognostic signature based on MRGs for LUAD and validated the performance of the model, which may provide a promising tool for the diagnosis, individualized immuno-/chemotherapeutic strategies and prognosis in patients with LUAD. [ABSTRACT FROM AUTHOR]

Background: Dysregulation of the balance between proliferation and apoptosis is the basis for human hepatocarcinogenesis. In many malignant tumors, such as hepatocellular carcinoma (HCC), there is a correlation between apoptotic dysregulation and poor prognosis. However, the prognostic values of apoptosis-related genes (ARGs) in HCC have not been elucidated. Methods: To screen for differentially expressed ARGs, the expression levels of 161 ARGs from The Cancer Genome Atlas (TCGA) database (https://cancergenome.nih.gov/) were analyzed. Gene Ontology (GO) enrichment and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed to evaluate the underlying molecular mechanisms of differentially expressed ARGs in HCC. The prognostic values of ARGs were established using Cox regression, and subsequently, a prognostic risk model for scoring patients was developed. Kaplan–Meier (K-M) and receiver operating characteristic (ROC) curves were plotted to determine the prognostic value of the model. Results: Compared with normal tissues, 43 highly upregulated and 8 downregulated ARGs in HCC tissues were screened. GO analysis results revealed that these 51 genes are indeed related to the apoptosis function. KEGG analysis revealed that these 51 genes were correlated with MAPK, P53, TNF, and PI3K-AKT signaling pathways, while Cox regression revealed that 5 ARGs (PPP2R5B, SQSTM1, TOP2A, BMF, and LGALS3) were associated with prognosis and were, therefore, obtained to develop the prognostic model. Based on the median risk scores, patients were categorized into high-risk and low-risk groups. Patients in the low-risk groups exhibited significantly elevated 2-year or 5-year survival probabilities (p < 0.0001). The risk model had a better clinical potency than the other clinical characteristics, with the area under the ROC curve (AUC = 0.741). The prognosis of HCC patients was established from a plotted nomogram. Conclusion: Based on the differential expression of ARGs, we established a novel risk model for predicting HCC prognosis. This model can also be used to inform the individualized treatment of HCC patients. [ABSTRACT FROM AUTHOR]

Background: Human cell division cycle associated 8 (CDCA8) a key regulator of mitosis, has been described as a potential prognostic biomarker for a variety of cancers, such as breast, colon and lung cancers. We aimed to evaluate the potential role of CDCA8 expression in the prognosis of liver cancer by analysing data from The Cancer Genome Atlas (TCGA). Methods: The Wilcoxon rank-sum test was used to compare the difference in CDCA8 expression between liver cancer tissues and matched normal tissues. Then, we applied logistic regression and the Wilcoxon rank-sum test to identify the association between CDCA8 expression and clinicopathologic characteristics. Cox regression and the Kaplan–Meier method were used to examine the clinicopathologic features correlated with overall survival (OS) in patients from the TCGA. Gene set enrichment analysis (GSEA) was performed to explore possible mechanisms of CDCA8 according to the TCGA dataset. Results: CDCA8 expression was higher in liver cancer tissues than in matched normal tissues. Logistic regression and the Wilcoxon rank-sum test revealed that the increased level of CDCA8 expression in liver cancer tissues was notably related to T stage (OR = 1.64 for T1/2 vs. T3/4), clinical stage (OR = 1.66 for I/II vs. III/IV), histologic grade (OR = 6.71 for G1 vs. G4) and histological type (OR = 0.24 for cholangiocarcinoma [CHOL] vs. hepatocellular carcinoma [LIHC]) (all P-values < 0.05). Kaplan–Meier survival analysis indicated that high CDCA8 expression was related to a poor prognosis in liver cancer (P = 2.456 × 10−6). Univariate analysis showed that high CDCA8 expression was associated with poor OS in liver cancer patients, with a hazard ratio (HR) of 1.85 (95% confidence interval [CI]: 1.47–2.32; P = 1.16 × 10–7). Multivariate analysis showed that CDCA8 expression was independently correlated with OS (HR = 1.74; CI: 1.25–12.64; P = 1.27 × 10–5). GSEA revealed that the apoptosis, cell cycle, ErbB, MAPK, mTOR, Notch, p53 and TGF-β signaling pathways were differentially enriched in the CDCA8 high expression phenotype. Conclusions: High CDCA8 expression is a potential molecular predictor of a poor prognosis in liver cancer. [ABSTRACT FROM AUTHOR]

Background: Gastric cancer (GC) has a high mortality rate and is one of the most fatal malignant tumours. Male sex has been proven as an independent risk factor for GC. This study aimed to identify immune-related genes (IRGs) associated with the prognosis of male GC. Methods: RNA sequencing and clinical data were obtained from The Cancer Genome Atlas (TCGA) database. Differentially expressed IRGs between male GC and normal tissues were identified by integrated bioinformatics analysis. Univariate and multivariate Cox regression analyses were applied to screen survival-associated IRGs. Then, GC patients were separated into high- and low-risk groups based on the median risk score. Furthermore, a nomogram was constructed based on the TCGA dataset. The prognostic value of the risk signature model was evaluated by Kaplan-Meier curve, receiver operating characteristic (ROC), Harrell's concordance index and calibration curves. In addition, the gene expression dataset from the Gene Expression Omnibus (GEO) was also downloaded for external validation. The relative proportions of 22 types of infiltrating immune cells in each male GC sample were evaluated using CIBERSORT. Results: A total of 276 differentially expressed IRGs were screened, including 189 up-regulated and 87 down-regulated genes. Subsequently, a seven-IRGs signature (LCN12, CCL21, RNASE2, CGB5, NRG4, AGTR1 and NPR3) was identified to be significantly associated with the overall survival (OS) of male GC patients. Survival analysis indicated that patients in the high-risk group exhibited a poor clinical outcome. The results of multivariate analysis revealed that the risk score was an independent prognostic factor. The established nomogram could be used to evaluate the prognosis of individual male GC patients. Further analysis showed that the prognostic model had excellent predictive performance in both TCGA and validated cohorts. Besides, the results of tumour-infiltrating immune cell analysis indicated that the seven-IRGs signature could reflect the status of the tumour immune microenvironment. Conclusions: Our study developed a novel seven-IRGs risk signature for individualized survival prediction of male GC patients. [ABSTRACT FROM AUTHOR]

Background: Frequent mutations in the nuclear receptor binding SET domain protein 1 (NSD1) gene have been observed in head and neck squamous cell carcinomas (HNSCC). NSD1 encodes a histone 3 lysine-36 methyltransferase. NSD1 mutations are correlated with improved clinical outcomes and increased sensitivity to platinum-based chemotherapy agents in human papillomavirus-negative (HPV-) tumors, despite weak T-cell infiltration. However, the role of NSD1 and related family members NSD2 and NSD3 in human papillomavirus-positive (HPV+) HNSCC is unclear. Methods: Using data from over 500 HNSCC patients from The Cancer Genome Atlas (TCGA), we compared the relative level of mRNA expression of NSD1, NSD2, and NSD3 in HPV+ and HPV- HNSCC. Correlation analyses were performed between T-cell infiltration and the relative level of expression of NSD1, NSD2, and NSD3 mRNA in HPV+ and HPV- HNSCC. In addition, overall survival outcomes were compared for both the HPV+ and HPV- subsets of patients based on stratification by NSD1, NSD2, and NSD3 expression levels. Results: Expression levels of NSD1, NSD2 or NSD3 were not correlated with altered lymphocyte infiltration in HPV+ HNSCC. More importantly, low expression of NSD1, NSD2, or NSD3 correlated with significantly reduced overall patient survival in HPV+, but not HPV- HNSCC. Conclusion: These results starkly illustrate the contrast in molecular features between HPV+ and HPV- HNSCC tumors and suggest that NSD1, NSD2, and NSD3 expression levels should be further investigated as novel clinical metrics for improved prognostication and patient stratification in HPV+ HNSCC. [ABSTRACT FROM AUTHOR]

Background: Globally, endometrial cancer is the fourth most common malignant tumor in women and the number of women being diagnosed is increasing. Tumor progression is strongly related to the cell survival-promoting functions of autophagy. We explored the relationship between endometrial cancer prognoses and the expression of autophagy genes using human autophagy databases.Methods: The Cancer Genome Atlas was used to identify autophagy related genes (ARGs) that were differentially expressed in endometrial cancer tissue compared to healthy endometrial tissue. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes were referenced to identify important biological functions and signaling pathways related to these differentially expressed ARGs. A prognostic model for endometrial cancer was constructed using univariate and multivariate Cox, and Least Absolute Shrinkage and Selection Operator regression analysis. Endometrial cancer patients were divided into high- and low-risk groups according to risk scores. Survival and receiver operating characteristic (ROC) curves were plotted for these patients to assess the accuracy of the prognostic model. Using immunohistochemistry the protein levels of the genes associated with risk were assessed.Results: We determined 37 ARGs were differentially expressed between endometrial cancer and healthy tissues. These genes were enriched in the biological processes and signaling pathways related to autophagy. Four ARGs (CDKN2A, PTK6, ERBB2 and BIRC5) were selected to establish a prognostic model of endometrial cancer. Kaplan-Meier survival analysis suggested that high-risk groups have significantly shorter survival times than low-risk groups. The area under the ROC curve indicated that the prognostic model for survival prediction was relatively accurate. Immunohistochemistry suggested that among the four ARGs the protein levels of CDKN2A, PTK6, ERBB2, and BIRC5 were higher in endometrial cancer than healthy endometrial tissue.Conclusions: Our prognostic model assessing four ARGs (CDKN2A, PTK6, ERBB2, and BIRC5) suggested their potential as independent predictive biomarkers and therapeutic targets for endometrial cancer. [ABSTRACT FROM AUTHOR]

Background: Exploring novel and sensitive targets is urgent due to the high morbidity of endometrial cancer (EC). The purpose of our study was to explore the transcription factors and immune-related genes in EC and further identify immune-based lncRNA signature as biomarker for predicting survival prognosis. Methods: Transcription factors, aberrantly expressed immune-related genes and immune-related lncRNAs were explored through bioinformatics analysis. Cox regression and the least absolute shrinkage and selection operator (LASSO) analysis were conducted to identify the immune and overall survival (OS) related lncRNAs. The accuracy of model was evaluated by Kaplan–Meier method and receiver operating characteristic (ROC) analysis, and the independent prognostic indicator was identified with Cox analysis. Quantitative real-time polymerase chain reaction (qRT-PCR) were conducted to detect the accuracy of our results. Results: A network of 29 transcription factors and 17 immune-related genes was constructed. Furthermore, four immune-prognosis-related lncRNAs were screened out. Kaplan–Meier survival analysis and time-dependent ROC analysis revealed a satisfactory predictive potential of the 4-lncRNA model. Consistency was achieved among the results from the training set, testing set and entire cohort. The distributed patterns between the high- and low-risk groups could be distinguished in principal component analysis. Comparisons of the risk score and clinical factors confirmed the four-lncRNA-based signature as an independent prognostic indicator. Last, the reliability of the results was verified by qRT-PCR in 29 cases of endometrial carcinoma and in cells. Conclusions: Overall, our study constructed a network of transcription factors and immune-related genes and explored a four immune-related lncRNA signature that could serve as a novel potential biomarker of EC. [ABSTRACT FROM AUTHOR]

Background: Although there are many studies on the characteristics of miRNA-mRNA interactions using miRNA and mRNA sequencing data, the complexity of the change of the correlation coefficients and expression values of the miRNA-mRNA pairs between tumor and normal samples is still not resolved, and this hinders the potential clinical applications. There is an urgent need to develop innovative methodologies and tools that can characterize and visualize functional consequences of cancer risk gene and miRNA pairs while analyzing the tumor and normal samples simultaneously. Results: We developed an innovative bioinformatics tool for visualizing functional annotation of miRNA-mRNA pairs in a network, known as MMiRNA-Viewer2. The tool takes mRNA and miRNA interaction pairs and visualizes mRNA and miRNA regulation network. Moreover, our MMiRNA-Viewer2 web server integrates and displays the mRNA and miRNA gene annotation information, signaling cascade pathways and direct cancer association between miRNAs and mRNAs. Functional annotation and gene regulatory information can be directly retrieved from our web server, which can help users quickly identify significant interaction sub-network and report possible disease or cancer association. The tool can identify pivotal miRNAs or mRNAs that contribute to the complexity of cancer, while engaging modern next-generation sequencing technology to analyze the tumor and normal samples concurrently. We compared our tools with other visualization tools. Conclusion: Our MMiRNA-Viewer2 serves as a multitasking platform in which users can identify significant interaction clusters and retrieve functional and cancer-associated information for miRNA-mRNA pairs between tumor and normal samples. Our tool is applicable across a range of diseases and cancers and has advantages over existing tools. [ABSTRACT FROM AUTHOR]

Background: Alternative splicing (AS) is a key factor in protein-coding gene diversity, and is associated with the development and progression of malignant tumours. However, the role of AS in cervical cancer is unclear. Methods: The AS data for cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC) were downloaded from The Cancer Genome Atlas (TCGA) SpliceSeq website. Few prognostic AS events were identified through univariate Cox analysis. We further identified the prognostic prediction models of the seven subtypes of AS events and assessed their predictive power. We constructed a clinical prediction model through global analysis of prognostic AS events and established a nomogram using the risk score calculated from the prognostic model and relevant clinical information. Unsupervised cluster analysis was used to explore the relationship between prognostic AS events in the model and clinical features. Results: A total of 2860 prognostic AS events in cervical cancer were identified. The best predictive effect was shown by a single alternate acceptor subtype with an area under the curve of 0.96. Our clinical prognostic model included a nine-AS event signature, and the c-index of the predicted nomogram model was 0.764. SNRPA and CCDC12 were hub genes for prognosis-associated splicing factors. Unsupervised cluster analysis through the nine prognostic AS events revealed three clusters with different survival patterns. Conclusions: AS events affect the prognosis and biological progression of cervical cancer. The identified prognostic AS events and splicing regulatory networks can increase our understanding of the underlying mechanisms of cervical cancer, providing new therapeutic strategies. [ABSTRACT FROM AUTHOR]

Background: The prognosis of hepatocellular carcinoma (HCC) patients remains poor. Identifying prognostic markers to stratify HCC patients might help to improve their outcomes. Methods: Six gene expression profiles (GSE121248, GSE84402, GSE65372, GSE51401, GSE45267 and GSE14520) were obtained for differentially expressed genes (DEGs) analysis between HCC tissues and non-tumor tissues. To identify the prognostic genes and establish risk score model, univariable Cox regression survival analysis and Lasso-penalized Cox regression analysis were performed based on the integrated DEGs by robust rank aggregation method. Then Kaplan–Meier and time-dependent receiver operating characteristic (ROC) curves were generated to validate the prognostic performance of risk score in training datasets and validation datasets. Multivariable Cox regression analysis was used to identify independent prognostic factors in liver cancer. A prognostic nomogram was constructed based on The Cancer Genome Atlas (TCGA) dataset. Finally, the correlation between DNA methylation and prognosis-related genes was analyzed. Results: A twelve-gene signature including SPP1, KIF20A, HMMR, TPX2, LAPTM4B, TTK, MAGEA6, ANX10, LECT2, CYP2C9, RDH16 and LCAT was identified, and risk score was calculated by corresponding coefficients. The risk score model showed a strong diagnosis performance to distinguish HCC from normal samples. The HCC patients were stratified into high-risk and low-risk group based on the cutoff value of risk score. The Kaplan–Meier survival curves revealed significantly favorable overall survival in groups with lower risk score (P < 0.0001). Time-dependent ROC analysis showed well prognostic performance of the twelve-gene signature, which was comparable or superior to AJCC stage at predicting 1-, 3-, and 5-year overall survival. In addition, the twelve-gene signature was independent with other clinical factors and performed better in predicting overall survival after combining with age and AJCC stage by nomogram. Moreover, most of the prognostic twelve genes were negatively correlated with DNA methylation in HCC tissues, which SPP1 and LCAT were identified as the DNA methylation-driven genes. Conclusions: We identified a twelve-gene signature as a robust marker with great potential for clinical application in risk stratification and overall survival prediction in HCC patients. [ABSTRACT FROM AUTHOR]

Background: Aberrant alternative splicing (AS) is implicated in biological processes of cancer. This study aims to reveal prognostic AS events and signatures that may serve as prognostic predictors for head and neck squamous cell carcinoma (HNSCC). Methods: Prognostic AS events in HNSCC were identified by univariate COX analysis. Prognostic signatures comprising prognostic AS events were constructed for prognosis prediction in patients with HNSCC. The correlation between the percent spliced in (PSI) values of AS events and the expression of splicing factors (SFs) was analyzed by Pearson correlation analysis. Gene functional annotation analysis was performed to reveal pathways in which prognostic AS is enriched. Results: A total of 27,611 AS events in 15,873 genes were observed, and there were 3433 AS events in 2624 genes significantly associated with overall survival (OS) for HNSCC. Moreover, we found that AS prognostic signatures could accurately predict HNSCC prognosis. SF-AS regulatory networks were constructed according to the correlation between PSI values of AS events and the expression levels of SFs. Conclusions: Our study identified prognostic AS events and signatures. Furthermore, it established SF-AS networks in HNSCC that were valuable in predicting the prognosis of patients with HNSCC and elucidating the regulatory mechanisms underlying AS in HNSCC. [ABSTRACT FROM AUTHOR]

Background: A major impediment in the treatment of ovarian cancer is the relapse of chemotherapy-resistant tumors, which occurs in approximately 25% of patients. A better understanding of the biological mechanisms underlying chemotherapy resistance will improve treatment efficacy through genetic testing and novel therapies.Methods: Using data from high-grade serous ovarian carcinoma (HGSOC) patients in the Cancer Genome Atlas (TCGA), we classified those who remained progression-free for 12 months following platinum-taxane combination chemotherapy as "chemo-sensitive" (N = 160) and those who had recurrence within 6 months as "chemo-resistant" (N = 110). Univariate and multivariate analysis of expression microarray data were used to identify differentially expressed genes and co-expression gene networks associated with chemotherapy response. Moreover, we integrated genomics data to determine expression quantitative trait loci (eQTL).Results: Differential expression of the Valosin-containing protein (VCP) gene and five co-expression gene networks were significantly associated with chemotherapy response in HGSOC. VCP and the most significant co-expression network module contribute to protein processing in the endoplasmic reticulum, which has been implicated in chemotherapy response. Both univariate and multivariate analysis findings were successfully replicated in an independent ovarian cancer cohort. Furthermore, we identified 192 cis-eQTLs associated with the expression of network genes and 4 cis-eQTLs associated with BRCA2 expression.Conclusion: This study implicates both known and novel genes as well as biological processes underlying response to platinum-taxane-based chemotherapy among HGSOC patients. [ABSTRACT FROM AUTHOR]

Background: Precise prediction of cancer types is vital for cancer diagnosis and therapy. Through a predictive model, important cancer marker genes can be inferred. Several studies have attempted to build machine learning models for this task however none has taken into consideration the effects of tissue of origin that can potentially bias the identification of cancer markers. Results: In this paper, we introduced several Convolutional Neural Network (CNN) models that take unstructured gene expression inputs to classify tumor and non-tumor samples into their designated cancer types or as normal. Based on different designs of gene embeddings and convolution schemes, we implemented three CNN models: 1D-CNN, 2D-Vanilla-CNN, and 2D-Hybrid-CNN. The models were trained and tested on gene expression profiles from combined 10,340 samples of 33 cancer types and 713 matched normal tissues of The Cancer Genome Atlas (TCGA). Our models achieved excellent prediction accuracies (93.9–95.0%) among 34 classes (33 cancers and normal). Furthermore, we interpreted one of the models, 1D-CNN model, with a guided saliency technique and identified a total of 2090 cancer markers (108 per class on average). The concordance of differential expression of these markers between the cancer type they represent and others is confirmed. In breast cancer, for instance, our model identified well-known markers, such as GATA3 and ESR1. Finally, we extended the 1D-CNN model for the prediction of breast cancer subtypes and achieved an average accuracy of 88.42% among 5 subtypes. The codes can be found at https://github.com/chenlabgccri/CancerTypePrediction. Conclusions: Here we present novel CNN designs for accurate and simultaneous cancer/normal and cancer types prediction based on gene expression profiles, and unique model interpretation scheme to elucidate biologically relevance of cancer marker genes after eliminating the effects of tissue-of-origin. The proposed model has light hyperparameters to be trained and thus can be easily adapted to facilitate cancer diagnosis in the future. [ABSTRACT FROM AUTHOR]

Background: Given that abnormal autophagy is involved in the pathogenesis of cancers, we sought to explore the potential value of autophagy-associated genes in lung adenocarcinoma (LUAD).Methods: RNA sequencing and clinical data on tumour and normal samples were acquired from The Cancer Genome Atlas (TCGA) database and randomly assigned to training and testing groups. Differentially expressed autophagy-associated genes (AAGs) were screened. Within the training group, Cox regression and Lasso regression analyses were conducted to screen five prognostic AAGs, which were used to develop a model. Kaplan-Meier (KM) and receiver operating characteristic (ROC) curves were plotted to determine the performance of the model in both groups. Immunohistochemistry was used to demonstrate the differential expression of AAGs in tumour and normal tissues at the protein level. Gene Ontology (GO) functional annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were utilized to further elucidate the roles of AAGs in LUAD.Results: The data from the TCGA database included 497 tumour and 54 normal samples, within which 30 differentially expressed AAGs were screened. Using Cox regression and Lasso regression analyses for the training group, 5 prognostic AAGs were identified and the prognostic model was constructed. Patients with low risk had better overall survival (OS) in the training group (3-year OS, 73.0% vs 48.0%; 5-year OS, 45.0% vs 33.8%; P = 1.305E-04) and in the testing group (3-year OS, 66.8% vs 41.2%; 5-year OS, 31.7% vs 25.8%; P = 1.027E-03). The areas under the ROC curves (AUC) were significant for both the training and testing groups (3-year AUC, 0.810 vs 0.894; 5-year AUC, 0.792 vs 0.749).Conclusions: We developed a survival model for LUAD and validated the performance of the model, which may provide superior outcomes for the patients. [ABSTRACT FROM AUTHOR]

Background: The study of high-throughput genomic profiles from a pharmacogenomics viewpoint has provided unprecedented insights into the oncogenic features modulating drug response. A recent study screened for the response of a thousand human cancer cell lines to a wide collection of anti-cancer drugs and illuminated the link between cellular genotypes and vulnerability. However, due to essential differences between cell lines and tumors, to date the translation into predicting drug response in tumors remains challenging. Recently, advances in deep learning have revolutionized bioinformatics and introduced new techniques to the integration of genomic data. Its application on pharmacogenomics may fill the gap between genomics and drug response and improve the prediction of drug response in tumors. Results: We proposed a deep learning model to predict drug response (DeepDR) based on mutation and expression profiles of a cancer cell or a tumor. The model contains three deep neural networks (DNNs), i) a mutation encoder pre-trained using a large pan-cancer dataset (The Cancer Genome Atlas; TCGA) to abstract core representations of high-dimension mutation data, ii) a pre-trained expression encoder, and iii) a drug response predictor network integrating the first two subnetworks. Given a pair of mutation and expression profiles, the model predicts IC50 values of 265 drugs. We trained and tested the model on a dataset of 622 cancer cell lines and achieved an overall prediction performance of mean squared error at 1.96 (log-scale IC50 values). The performance was superior in prediction error or stability than two classical methods (linear regression and support vector machine) and four analog DNN models of DeepDR, including DNNs built without TCGA pre-training, partly replaced by principal components, and built on individual types of input data. We then applied the model to predict drug response of 9059 tumors of 33 cancer types. Using per-cancer and pan-cancer settings, the model predicted both known, including EGFR inhibitors in non-small cell lung cancer and tamoxifen in ER+ breast cancer, and novel drug targets, such as vinorelbine for TTN-mutated tumors. The comprehensive analysis further revealed the molecular mechanisms underlying the resistance to a chemotherapeutic drug docetaxel in a pan-cancer setting and the anti-cancer potential of a novel agent, CX-5461, in treating gliomas and hematopoietic malignancies. Conclusions: Here we present, as far as we know, the first DNN model to translate pharmacogenomics features identified from in vitro drug screening to predict the response of tumors. The results covered both well-studied and novel mechanisms of drug resistance and drug targets. Our model and findings improve the prediction of drug response and the identification of novel therapeutic options. [ABSTRACT FROM AUTHOR]

Background: Alternative polyadenylation (APA) results in messenger RNA molecules with different 3′ untranslated regions (3’ UTRs), affecting the molecules’ stability, localization, and translation. APA is pervasive and implicated in cancer. Earlier reports on APA focused on 3’ UTR length modifications and commonly characterized APA events as 3’ UTR shortening or lengthening. However, such characterization oversimplifies the processing of 3′ ends of transcripts and fails to adequately describe the various scenarios we observe. Results: We built a cloud-based targeted de novo transcript assembly and analysis pipeline that incorporates our previously developed cleavage site prediction tool, KLEAT. We applied this pipeline to elucidate the APA profiles of 114 genes in 9939 tumor and 729 tissue normal samples from The Cancer Genome Atlas (TCGA). The full set of 10,668 RNA-Seq samples from 33 cancer types has not been utilized by previous APA studies. By comparing the frequencies of predicted cleavage sites between normal and tumor sample groups, we identified 77 events (i.e. gene-cancer type pairs) of tumor-specific APA regulation in 13 cancer types; for 15 genes, such regulation is recurrent across multiple cancers. Our results also support a previous report showing the 3’ UTR shortening of FGF2 in multiple cancers. However, over half of the events we identified display complex changes to 3’ UTR length that resist simple classification like shortening or lengthening. Conclusions: Recurrent tumor-specific regulation of APA is widespread in cancer. However, the regulation pattern that we observed in TCGA RNA-seq data cannot be described as straightforward 3’ UTR shortening or lengthening. Continued investigation into this complex, nuanced regulatory landscape will provide further insight into its role in tumor formation and development. [ABSTRACT FROM AUTHOR]

Background: Colorectal cancer (CRC) is 3rd most commonly diagnosed cancer in males and the second in females. PD-1/PD-L1 axis, as an immune checkpoint, is up-regulated in many tumors and their microenvironment. However, the prognostic value of PD-1/PD-L1 in CRC remains unclear. Methods: The Cancer Genome Atlas (TCGA) database (N = 356) and Fudan University Shanghai Cancer Center (FUSCC) cohort of patients (N = 276) were adopted to analyze the prognostic value of PD-L1 in colorectal tumor cells (TCs) and of PD-1 in tumor infiltrating cells (TILs) for CRC. Subgroup analyses were conducted in FUSCC cohort according to patients' status of mismatch repair. Results: In TCGA cohort, the cut-off values of PD-1 and PD-L1 expression were determined by X-tile program, which were 4.40 and 2.92, respectively. Kaplan-Meier analysis indicated that higher PD-1 and PD-L1 expressions correlated with better OS (P = 0.032 and P = 0.002, respectively). In FUSCC cohort, expressions of PD-1 on TILs and PD-L1 on TCs were analyzed separately by immunohistochemistry (IHC) staining based on a TMA sample (N = 276) and revealed that both TILs-PD-1 and TCs-PD-L1 were associated with OS (P = 0.006 and P = 0.002, respectively) and DFS (P = 0.025 and P = 0.004, respectively) of CRC patients. Multivariate Cox regression analysis indicated TILs- PD-1 was an independent prognostic factor both for OS and DFS of CRC patients (P < 0.05). Subgroup analyses showed that TILs-PD-1 was an independent prognostic factor for both OS and DFS in CRC patients in MSSproficient subgroup (P < 0.05), while neither of them correlated with OS or DFS in MSS-deficient subgroup (P > 0.05). Conclusions: Higher expressions of PD-1 and PD-L1 correlates with better prognosis of CRC patients. TILs-PD-1 is an independent prognostic factor for OS and DFS of CRC patients, especially for MMR-proficient subgroup. [ABSTRACT FROM AUTHOR]

Background: Technological advances enable the cost-effective acquisition of Multi-Modal Data Sets (MMDS) composed of measurements for multiple, high-dimensional data types obtained from a common set of bio-samples. The joint analysis of the data matrices associated with the different data types of a MMDS should provide a more focused view of the biology underlying complex diseases such as cancer that would not be apparent from the analysis of a single data type alone. As multi-modal data rapidly accumulate in research laboratories and public databases such as The Cancer Genome Atlas (TCGA), the translation of such data into clinically actionable knowledge has been slowed by the lack of computational tools capable of analyzing MMDSs. Here, we describe the Joint Analysis of Many Matrices by ITeration (JAMMIT) algorithm that jointly analyzes the data matrices of a MMDS using sparse matrix approximations of rank-1. Methods: The JAMMIT algorithm jointly approximates an arbitrary number of data matrices by rank-1 outer-products composed of "sparse" left-singular vectors (eigen-arrays) that are unique to each matrix and a right-singular vector (eigen-signal) that is common to all the matrices. The non-zero coefficients of the eigen-arrays identify small subsets of variables for each data type (i.e., signatures) that in aggregate, or individually, best explain a dominant eigen-signal defined on the columns of the data matrices. The approximation is specified by a single "sparsity" parameter that is selected based on false discovery rate estimated by permutation testing. Multiple signals of interest in a given MDDS are sequentially detected and modeled by iterating JAMMIT on "residual" data matrices that result from a given sparse approximation. Results: We show that JAMMIT outperforms other joint analysis algorithms in the detection of multiple signatures embedded in simulated MDDS. On real multimodal data for ovarian and liver cancer we show that JAMMIT identified multi-modal signatures that were clinically informative and enriched for cancer-related biology. Conclusions: Sparse matrix approximations of rank-1 provide a simple yet effective means of jointly reducing multiple, big data types to a small subset of variables that characterize important clinical and/or biological attributes of the bio-samples from which the data were acquired. [ABSTRACT FROM AUTHOR]

Background Over the past years, tremendous efforts have been made to elucidate the molecular basis of the initiation and progression of ovarian cancer. However, most existing studies have been focused on individual genes or a single type of data, which may lack the power to detect the complex mechanism of cancer formation by overlooking the interactions of different genetic and epigenetic factors. Results We propose an integrative framework to identify genetic and epigenetic features related to ovarian cancer and to quantify the causal relationships among these features using a probabilistic graphical model based on the Cancer Genome Atlas (TCGA) data. In the feature selection, we first defined a set of seed genes by including 48 candidate tumor suppressors or oncogenes and an additional 20 ovarian cancer related genes reported in the literature. The seed genes were then fed into a stepwise correlation-based selector to identify 271 additional features including 177 genes, 82 copy number variation sites, 11 methylation sites and 1 somatic mutation (at gene TP53). We built a Bayesian network model with a logit link function to quantify the causal relationship among these features and discovered a set of 13 hub genes including ARID1A, C19orf53, CSKN2A1 and COL5A2. The directed graph revealed many potential genetic pathways, some of which confirmed the existing results in the literature. Clustering analysis further suggested four gene clusters, three of which correspond to well-defined cellular processes including cell division, tumor invasion and mitochondrial system. In addition, two genes related to glycoprotein synthesis, PSG11 and GALNT10, were found highly predictive for the overall survival time of ovarian cancer patients. Conclusions The proposed framework is effective in identifying possible important genetic and epigenetic features that are related to complex cancer diseases. The constructed Bayesian network has identified some new genetic/epigenetic pathways, which may shed new light into the molecular mechanisms of ovarian cancer. [ABSTRACT FROM AUTHOR]

Background: Aberrant DNA methylation is frequently observed in breast cancer. However, the relationship between methylation patterns and the heterogeneity of breast cancer has not been comprehensively characterized.Methods: Whole-genome DNA methylation analysis using Illumina Infinium HumanMethylation450 BeadChip arrays was performed on 188 human breast tumors. Unsupervised bootstrap consensus clustering was performed to identify DNA methylation epigenetic subgroups (epitypes). The Cancer Genome Atlas data, including methylation profiles of 669 human breast tumors, was used for validation. The identified epitypes were characterized by integration with publicly available genome-wide data, including gene expression levels, DNA copy numbers, whole-exome sequencing data, and chromatin states.Results: We identified seven breast cancer epitypes. One epitype was distinctly associated with basal-like tumors and with BRCA1 mutations, one epitype contained a subset of ERBB2-amplified tumors characterized by multiple additional amplifications and the most complex genomes, and one epitype displayed a methylation profile similar to normal epithelial cells. Luminal tumors were stratified into the remaining four epitypes, with differences in promoter hypermethylation, global hypomethylation, proliferative rates, and genomic instability. Specific hyper- and hypomethylation across the basal-like epitype was rare. However, we observed that the candidate genomic instability drivers BRCA1 and HORMAD1 displayed aberrant methylation linked to gene expression levels in some basal-like tumors. Hypomethylation in luminal tumors was associated with DNA repeats and subtelomeric regions. We observed two dominant patterns of aberrant methylation in breast cancer. One pattern, constitutively methylated in both basal-like and luminal breast cancer, was linked to genes with promoters in a Polycomb-repressed state in normal epithelial cells and displayed no correlation with gene expression levels. The second pattern correlated with gene expression levels and was associated with methylation in luminal tumors and genes with active promoters in normal epithelial cells.Conclusions: Our results suggest that hypermethylation patterns across basal-like breast cancer may have limited influence on tumor progression and instead reflect the repressed chromatin state of the tissue of origin. On the contrary, hypermethylation patterns specific to luminal breast cancer influence gene expression, may contribute to tumor progression, and may present an actionable epigenetic alteration in a subset of luminal breast cancers. [ABSTRACT FROM AUTHOR]