A human functional protein interaction network and its application to cancer data analysis

Pathogenic gene prediction based on network embedding

In disease research, the study of gene-disease correlation has always been an important topic. With the emergence of large-scale connected data sets in biology, we use known correlations between the entities, which may be from different sets, to build a biological heterogeneous network and propose a new network embedded representation algorithm to calculate the correlation between disease and genes, using the correlation score to predict pathogenic genes. Then, we conduct several experiments to compare our method to other state-of-the-art methods. The results reveal that our method achieves better performance than the traditional methods.

Transcriptomic Profiling of Human Pluripotent Stem Cell-derived Retinal Pigment Epithelium over Time

Human pluripotent stem cell (hPSC)-derived progenies are immature versions of cells, presenting a potential limitation to the accurate modelling of diseases associated with maturity or age. Hence, it is important to characterise how closely cells used in culture resemble their native counterparts. In order to select appropriate time points of retinal pigment epithelium (RPE) cultures that reflect native counterparts, we characterised the transcriptomic profiles of the hPSC-derived RPE cells from 1- and 12-month cultures. We differentiated the human embryonic stem cell line H9 into RPE cells, performed single-cell RNA-sequencing of a total of 16,576 cells to assess the molecular changes of the RPE cells across these two culture time points. Our results indicate the stability of the RPE transcriptomic signature, with no evidence of an epithelial–mesenchymal transition, and with the maturing populations of the RPE observed with time in culture. Assessment of Gene Ontology pathways revealed that as the cultures age, RPE cells upregulate expression of genes involved in metal binding and antioxidant functions. This might reflect an increased ability to handle oxidative stress as cells mature. Comparison with native human RPE data confirms a maturing transcriptional profile of RPE cells in culture. These results suggest that long-term in vitro culture of RPE cells allows the modelling of specific phenotypes observed in native mature tissues. Our work highlights the transcriptional landscape of hPSC-derived RPE cells as they age in culture, which provides a reference for native and patient samples to be benchmarked against.

Computational identification of biomarker genes for lung cancer considering treatment and non-treatment studies

BACKGROUND

Lung cancer is the number one cancer killer in the world with more than 142,670 deaths estimated in the United States alone in the year 2019. Consequently, there is an overreaching need to identify the key biomarkers for lung cancer. The aim of this study is to computationally identify biomarker genes for lung cancer that can aid in its diagnosis and treatment. The gene expression profiles of two different types of studies, namely non-treatment and treatment, are considered for discovering biomarker genes. In non-treatment studies healthy samples are control and cancer samples are cases. Whereas, in treatment studies, controls are cancer cell lines without treatment and cases are cancer cell lines with treatment.

RESULTS

The Differentially Expressed Genes (DEGs) for lung cancer were isolated from Gene Expression Omnibus (GEO) database using R software tool GEO2R. A total of 407 DEGs (254 upregulated and 153 downregulated) from non-treatment studies and 547 DEGs (133 upregulated and 414 downregulated) from treatment studies were isolated. Two Cytoscape apps, namely, CytoHubba and MCODE, were used for identifying biomarker genes from functional networks developed using DEG genes. This study discovered two distinct sets of biomarker genes - one from non-treatment studies and the other from treatment studies, each set containing 16 genes. Survival analysis results show that most non-treatment biomarker genes have prognostic capability by indicating low-expression groups have higher chance of survival compare to high-expression groups. Whereas, most treatment biomarkers have prognostic capability by indicating high-expression groups have higher chance of survival compare to low-expression groups.

CONCLUSION

A computational framework is developed to identify biomarker genes for lung cancer using gene expression profiles. Two different types of studies - non-treatment and treatment - are considered for experiment. Most of the biomarker genes from non-treatment studies are part of mitosis and play vital role in DNA repair and cell-cycle regulation. Whereas, most of the biomarker genes from treatment studies are associated to ubiquitination and cellular response to stress. This study discovered a list of biomarkers, which would help experimental scientists to design a lab experiment for further exploration of detail dynamics of lung cancer development.

A Combined Analysis of Genetically Correlated Traits Identifies Genes and Brain Regions for Insomnia

AIMS

Previous studies have inferred that there is a strong genetic component in insomnia. However, the etiology of insomnia is still unclear. This study systematically analyzed multiple genome-wide association study (GWAS) data sets with core human pathways and functional networks to detect potential gene pathways and networks associated with insomnia.

METHODS

We used a novel method, multitrait analysis of genome-wide association studies (MTAG), to combine 3 large GWASs of insomnia symptoms/complaints and sleep duration. The i-Gsea4GwasV2 and Reactome FI programs were used to analyze data from the result of MTAG analysis and the nominally significant pathways, respectively.

RESULTS

Through analyzing data sets using the MTAG program, our sample size increased from 113,006 subjects to 163,188 subjects. A total of 17 of 1,816 Reactome pathways were identified and showed to be associated with insomnia. We further revealed 11 interconnected functional and topologically interacting clusters (Clusters 0 to 10) that were associated with insomnia. Based on the brain transcriptome data, it was found that the genes in Cluster 4 were enriched for the transcriptional coexpression profile in the prenatal dorsolateral prefrontal cortex (P = 7 × 10-5), inferolateral temporal cortex (P = 0.02), medial prefrontal cortex (P < 1 × 10-5), and amygdala (P < 1 × 10-5), and detected RPA2, ORC6, PIAS3, and PRIM2 as core nodes in these 4 brain regions.

CONCLUSIONS

The findings provided new genes, pathways, and brain regions to understand the pathology of insomnia.

Macroscopic somatic clonal expansion in morphologically normal human urothelium

Knowledge of somatic mutation accumulation in normal cells, which is essential for understanding cancer development and evolution, remains largely lacking. In this study, we investigated somatic clonal events in morphologically normal human urothelium (MNU; epithelium lining the bladder and ureter) and identified macroscopic clonal expansions. Aristolochic acid (AA), a natural herb-derived compound, was a major mutagenic driving factor in MNU. AA drastically accelerates mutation accumulation and enhances clonal expansion. Mutations in MNU were widely observed in chromatin remodeling genes such as KMT2D and KDM6A but rarely in TP53, PIK3CA, and FGFR3 KMT2D mutations were found to be common in urothelial cells, regardless of whether the cells experience exogenous mutagen exposure. Copy number alterations were rare and largely confined to small-scale regions, along with copy-neutral loss of heterozygosity. Single AA-associated clones in MNU expanded to a scale of several square centimeters in size.

Revealing Prognosis-Related Pathways at the Individual Level by a Comprehensive Analysis of Different Cancer Transcription Data

Identifying perturbed pathways at an individual level is important to discover the causes of cancer and develop individualized custom therapeutic strategies. Though prognostic gene lists have had success in prognosis prediction, using single genes that are related to the relevant system or specific network cannot fully reveal the process of tumorigenesis. We hypothesize that in individual samples, the disruption of transcription homeostasis can influence the occurrence, development, and metastasis of tumors and has implications for patient survival outcomes. Here, we introduced the individual-level pathway score, which can measure the correlation perturbation of the pathways in a single sample well. We applied this method to the expression data of 16 different cancer types from The Cancer Genome Atlas (TCGA) database. Our results indicate that different cancer types as well as their tumor-adjacent tissues can be clearly distinguished by the individual-level pathway score. Additionally, we found that there was strong heterogeneity among different cancer types and the percentage of perturbed pathways as well as the perturbation proportions of tumor samples in each pathway were significantly different. Finally, the prognosis-related pathways of different cancer types were obtained by survival analysis. We demonstrated that the individual-level pathway score (iPS) is capable of classifying cancer types and identifying some key prognosis-related pathways.

Machine learning techniques for sequence-based prediction of viral-host interactions between SARS-CoV-2 and human proteins

BACKGROUND

COVID-19 (Coronavirus Disease-19), a disease caused by the SARS-CoV-2 virus, has been declared as a pandemic by the World Health Organization on March 11, 2020. Over 15 million people have already been affected worldwide by COVID-19, resulting in more than 0.6 million deaths. Protein-protein interactions (PPIs) play a key role in the cellular process of SARS-CoV-2 virus infection in the human body. Recently a study has reported some SARS-CoV-2 proteins that interact with several human proteins while many potential interactions remain to be identified.

METHOD

In this article, various machine learning models are built to predict the PPIs between the virus and human proteins that are further validated using biological experiments. The classification models are prepared based on different sequence-based features of human proteins like amino acid composition, pseudo amino acid composition, and conjoint triad.

RESULT

We have built an ensemble voting classifier using SVMRadial, SVMPolynomial, and Random Forest technique that gives a greater accuracy, precision, specificity, recall, and F1 score compared to all other models used in the work. A total of 1326 potential human target proteins of SARS-CoV-2 have been predicted by the proposed ensemble model and validated using gene ontology and KEGG pathway enrichment analysis. Several repurposable drugs targeting the predicted interactions are also reported.

CONCLUSION

This study may encourage the identification of potential targets for more effective anti-COVID drug discovery.

A functionally defined high-density NRF2 interactome reveals new conditional regulators of ARE transactivation

NRF2 (NFE2L2) is a cytoprotective transcription factor associated with >60 human diseases, adverse drug reactions and therapeutic resistance. To provide insight into the complex regulation of NRF2 responses, 1962 predicted NRF2-partner interactions were systematically tested to generate an experimentally defined high-density human NRF2 interactome. Verification and conditional stratification of 46 new NRF2 partners was achieved by co-immunoprecipitation and the novel integration of quantitative data from dual luminescence-based co-immunoprecipitation (DULIP) assays and live-cell fluorescence cross-correlation spectroscopy (FCCS). The functional impact of new partners was then assessed in genetically edited loss-of-function (NRF2-/-) and disease-related gain-of-function (NRF2T80K and KEAP1-/-) cell-lines. Of the new partners investigated >77% (17/22) modified NRF2 responses, including partners that only exhibited effects under disease-related conditions. This experimentally defined binary NRF2 interactome provides a new vision of the complex molecular networks that govern the modulation and consequence of NRF2 activity in health and disease.

Identification of Survival-Associated Alternative Splicing Signatures in Lung Squamous Cell Carcinoma

Purpose: Alternative splicing (AS) is a post-transcriptional process that plays a significant role in enhancing the diversity of transcription and protein. Accumulating evidences have demonstrated that dysregulation of AS is associated with oncogenic processes. However, AS signature specifically in lung squamous cell carcinoma (LUSC) remains unknown. This study aimed to evaluate the prognostic values of AS events in LUSC patients.

Methods: The RNA-seq data, AS events data and corresponding clinical information were obtained from The Cancer Genome Atlas (TCGA) database. Univariate Cox regression analysis was performed to identify survival-related AS events and survival-related parent genes were subjected to Gene Ontology enrichment analysis and gene network analysis. The least absolute shrinkage and selection operator (LASSO) method and multivariate Cox regression analysis were used to construct prognostic prediction models, and their predictive values were assessed by Kaplan-Meier analysis and receiver operating characteristic (ROC) curves. Then a nomogram was established to predict the survival of LUSC patients. And the interaction network of splicing factors (SFs) and survival-related AS events was constructed by Spearman correlation analysis and visualized by Cytoscape.

Results: Totally, 467 LUSC patients were included in this study and 1,991 survival-related AS events within 1,433 genes were identified. SMAD4, FOS, POLR2L, and RNPS1 were the hub genes in the gene interaction network. Eight prognostic prediction models (seven types of AS and all AS) were constructed and all exhibited high efficiency in distinguishing good or poor survival of LUSC patients. The final integrated prediction model including all types of AS events exhibited the best prognostic power with the maximum AUC values of 0.778, 0.816, 0.814 in 1, 3, 5 years ROC curves, respectively. Meanwhile, the nomogram performed well in predicting the 1-, 3-, and 5-year survival of LUSC patients. In addition, the SF-AS regulatory network uncovered a significant correlation between SFs and survival-related AS events.

Conclusion: This is the first comprehensive study to analyze the role of AS events in LUSC specifically, which improves our understanding of the prognostic value of survival-related AS events for LUSC. And these survival-related AS events might serve as novel prognostic biomarkers and drug therapeutic targets for LUSC.

Integrative bioinformatics analysis of prognostic alternative splicing signatures in gastric cancer

Background

The potential prognostic value of alternative splicing (AS) variants and regulatory splicing factors in gastric carcinogenesis is unclear. We aimed to exploit the aberrant AS signatures and splicing factors involved in gastric cancer (GC) and to determine their prognostic predictive values.

Methods

We performed detailed data acquisition using the Cancer Genome Atlas project and profiled genome-wide AS signatures in a cohort of 190 patients with stomach adenocarcinoma (STAD). Prognostic prediction models and splicing correlation networks were assessed using an integrative bioinformatics analysis approach.

Results

We detected 1,308 overall survival (OS)-related AS signatures in 993 genes, most of which were favorable prognostic factors. Six splicing factors have been suggested to be dysregulated in GC, i.e., DHX15, PPP4R2, PRPF38B, RBM9, RBM15, and ILF3. Another notable finding was that most favorable prognosis AS events were positively correlated with expression of splicing factors, while a majority of poor survival prognostic AS genes were negatively associated with the expression of splicing factors.

Conclusions

To our knowledge, the current study provided the first comprehensive profiling of global modifications in the RNA splicing to identify survival associated AS signatures of GC specific genes. Our findings contribute to a better understanding of aberrant AS signatures and splicing factors in STAD, which can potentially be used as prognostic biomarkers and therapeutic targets for GC.

Tumor Microenvironment-Associated Immune-Related Genes for the Prognosis of Malignant Pleural Mesothelioma

Malignant pleural mesothelioma (MPM) is a rare but highly aggressive thoracic malignancy. ESTIMATE algorithm-derived immune scores are commonly used to quantify the immune and stromal components in tumors. Thus, this algorithm may help determine the tumor microenvironment (TME)-related gene expression profile associated with tumor immunity. This study aimed at mining public databases to determine a potential correlation between differentially expressed genes (DEGs) and survival in patients with MPM. We categorized patients from the Gene Expression Omnibus database according to their immune/stromal scores into high- and low-score groups. Functional enrichment analysis and the construction of protein-protein interaction networks showed that the DEGs identified were primarily involved in the TME. Furthermore, we validated these genes in an independent cohort of patients with MPM from The Cancer Genome Atlas database. DEG analysis showed that 29 DEGs were cancer driver genes. Subsequently, 14 TME-related genes, which have been previously neglected, were shown to exhibit significant prognostic potential in MPM. In conclusion, immune/stromal scores are novel predictors of a poor prognosis in patients with MPM. We identified DEGs that are involved in immunity against MPM and may contribute to patient survival. Owing to their potential as prognostic factors for MPM, these 14 TME-related genes need to be studied in detail in the future.

Determining the Molecular Background of Endometrial Receptivity in Adenomyosis

Background: Adenomyosis is a gynaecological condition with limited evidence of negative impact to endometrial receptivity. It is commonly associated with endometriosis, which has been shown to alter endometrial expression patterns. Therefore, the candidate genes identified in endometriosis could serve as a source to study endometrial function in adenomyosis. Methods: Transcripts/proteins associated with endometrial receptivity in women with adenomyosis or endometriosis and healthy women were obtained from publications and their nomenclature was adopted according to the HUGO Gene Nomenclature Committee (HGNC). Retrieved genes were analysed for enriched pathways using Cytoscape/Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) and Reactome tools to prioritise candidates for endometrial receptivity. These were used for validation on women with (n = 9) and without (n = 13) adenomyosis. Results: Functional enrichment analysis of 173, 42 and 151 genes associated with endometriosis, adenomyosis and healthy women, respectively, revealed signalling by interleukins and interleukin-4 and interleukin-13 signalling pathways, from which annotated LIF, JUNB, IL6, FOS, IL10 and SOCS3 were prioritised. Selected genes showed downregulated expression levels in adenomyosis compared to the control group, but without statistical significance. Conclusion: This is the first integrative study providing putative candidate genes and pathways characterising endometrial receptivity in women with adenomyosis in comparison to healthy women and women with endometriosis.

A comprehensive p75 neurotrophin receptor gene network and pathway analyses identifying new target genes

P75 neurotrophic receptor (p75NTR) is an important receptor for the role of neurotrophins in modulating brain plasticity and apoptosis. The current understanding of the role of p75NTR in cellular adaptation following pathological insults remains blurred, which makes p75NTR’s related signaling networks an interesting and challenging initial point of investigation. We identified p75NTR and related genes through extensive data mining of a PubMed literature search including published works related to p75NTR from the past 20 years. Bioinformatic network and pathway analyses of identified genes (n = 235) were performed using ReactomeFIViz in Cytoscape based on the highly reliable Reactome functional interaction network algorithm. This approach merges interactions extracted from human curated pathways with predicted interactions from machine learning. Genome-wide pathway analysis showed total of 16 enriched hierarchical clusters. A total of 278 enriched single pathways were also identified (p < 0.05, false discovery rate corrected). Gene network analyses showed multiple known and new targets in the p75NTR gene network. This study provides a comprehensive analysis and investigation into the current knowledge of p75NTR signaling networks and pathways. These results also identify several genes and their respective protein products as involved in the p75NTR network, which have not previously been clearly studied in this pathway. These results can be used to generate novel hypotheses to gain a greater understanding of p75NTR in acute brain injuries, neurodegenerative diseases and general response to cellular damage.

Identifying and ranking potential cancer drivers using representation learning on attributed network

Cancer can arise as a consequence of the accumulation of genomic alterations. Only a small part of driver mutations contributes to cancer development and progression. Hence, the identification of genes and alterations that serve as drivers for cancer development plays a critical role in drug design, cancer diagnoses and treatment. In this study, we propose a novel method to identify potential cancer drivers by using a Representation Learning method on Attributed Graphs (called RLAG). It is a first attempt to use both network structure and node attributes to learn feature representation for the genes in the network. Then it leverages these feature vectors to divide the genes into several subgroups. Finally, potential cancer driver genes are prioritized according to ranking scores that measure both genes' properties and their importance in the subgroups. We apply our method to predict driver genes for lung cancer, breast cancer and prostate cancer. The results show that our method outperforms the other three state-of-the-art methods in terms of Precision, Recall and F1-score values.

Approaches for the identification of driver mutations in cancer: A tutorial from a computational perspective

Cancer is a complex disease caused by the accumulation of genetic alterations during the individual's life. Such alterations are called genetic mutations and can be divided into two groups: (1) Passenger mutations, which are not responsible for cancer and (2) Driver mutations, which are significant for cancer and responsible for its initiation and progression. Cancer cells undergo a large number of mutations, of which most are passengers, and few are drivers. The identification of driver mutations is a key point and one of the biggest challenges in Cancer Genomics. Many computational methods for such a purpose have been developed in Cancer Bioinformatics. Such computational methods are complex and are usually described in a high level of abstraction. This tutorial details some classical computational methods, from a computational perspective, with the transcription in an algorithmic format towards an easy access by researchers.

Genome-wide analyses of the prognosis-related mRNA alternative splicing landscape and novel splicing factors based on large-scale low grade glioma cohort

Alternative splicing (AS) changes are considered to be critical in predicting treatment response. Our study aimed to investigate differential splicing patterns and to elucidate the role of splicing factor (SF) as prognostic markers of low-grade glioma (LGG). We downloaded RNA-seq data from a cohort of 516 LGG tumors in The Cancer Genome Atlas and analyzed independent prognostic factors using LASSO regression and Cox proportional regression to build a network based on the correlation between SF-related survival AS events. We collected 100 patients from our center for immunohistochemistry and analyzed survival using χ2 test and Cox and Kaplan-Meier analyses. A total of 9,616 AS events related to LGG were screened and identified as well as established related models. Through analyzing specific splicing patterns in LGG, we screened 16 genes to construct a prognostic model to stratify the risk of LGG patients. Validation revealed that the expression level of the prognostic model in LGG tissue was increased, and patients with high expression showed worse prognosis. In summary, we demonstrated the role of SFs and AS events in the progression of LGG, which may provide insights into the clinical significance and aid the future exploration of LGG-associated AS.

Genomic Tools for the Conservation and Genetic Improvement of a Highly Fragmented Breed-The Ramo Grande Cattle from the Azores

Simple Summary

Inbreeding control is a key concern in managing local endangered breeds, which often have developed unique adaptation features. Ramo Grande is a local cattle breed raised in the Azores archipelago under very harsh conditions, with a census of about 1300 cows dispersed by various islands. This fragmentation is a challenge when the goal is to keep inbreeding under control. Currently, panels of genetic markers are available which enable the assessment of inbreeding and the occurrence of previous bottlenecks in a population. These panels also allow the identification of genes associated with specific production traits, if reliable phenotypic information is available. We used a panel of genetic markers and estimated that the degree of inbreeding was approaching a level of concern, while some exotic gene inflow may have occurred in the past. We were able to identify genetic markers significantly associated with longevity, which reflects the ability of these cattle to remain productive under severe environmental conditions. Genetic markers were also identified as significantly associated with age at first calving and calf growth rate. The results indicate that genomic information can be used to control inbreeding and to implement genomic selection in Ramo Grande cattle to enhance adaptation and production traits.

Abstract

Ramo Grande is a local cattle breed raised in the archipelago of Azores, with a small and dispersed census, where inbreeding control is of utmost importance. A single nucleotide polymorphism (SNP) Beadchip array was used to assess inbreeding, by analysis of genomic regions harboring contiguous homozygous genotypes named runs of homozygosity (ROH), and to estimate past effective population size by analysis of linkage disequilibrium (LD). Genetic markers associated with production traits were also investigated, exploiting the unique genetic and adaptation features of this breed. A total of 639 ROH with length >4 Mb were identified, with mean length of 14.96 Mb. The mean genomic inbreeding was 0.09, and long segments of ROH were common, indicating recent inbred matings. The LD pattern indicates a large effective population size, suggesting the inflow of exotic germplasm in the past. The genome-wide association study identified novel markers significantly affecting longevity, age at first calving and direct genetic effects on calf weight. These results provide the first evidence of the association of longevity with genes related with DNA recognition and repair, and the association of age at first calving with aquaporin proteins, which are known to have a crucial role in reproduction.

Comprehensive Characterization of the Prognosis Value of Alternative Splicing Events in Acute Myeloid Leukemia

Increasing evidence have demonstrated that dysregulated alternative splicing (AS) events promoted tumor development and was correlated with worse prognosis in the context of certain malignancies. Nevertheless, a comprehensive examination of the prognosis role of AS events in acute myeloid leukemia (AML) has not yet been illuminated. In this study, univariate and multivariate Cox regression analysis were used to identify survival-related AS events and independent prognostic predictors. The interaction between splicing factors (SFs) and AS events was visualized by Cytoscape. A total of 3013 survival-associated AS events in 1977 genes were screened in 151 AML patients. Interestingly, the majority (2031 events) were revealed to be protective factors. Furthermore, the prediction models were constructed for each type of AS and all of them displayed good performance in predicting prognosis, considering their area under curve values of the receiver operating characteristic were all above 0.7. Notably, the splicing regulatory network displayed the underlying interaction networks between SFs and AS events. Taken together, our study demonstrated the survival-related AS events in AML and uncovered the possible association between SFs and prognostic AS events, which provide new prognostic biomarkers and aid to develop novel targets for AML therapy.

Epigenome-wide association study in healthy individuals identifies significant associations with DNA methylation and PBMC extract VEGF-A concentration

INTRODUCTION

Vascular endothelial growth factor A (VEGF-A) is a chemokine that induces proliferation and migration of vascular endothelial cells and is essential for both physiological and pathological angiogenesis. It is known for its high heritability (> 60%) and involvement in most common morbidities, which makes it a potentially interesting biomarker. Large GWAS studies have already assessed polymorphisms related to VEGF-A. However, no previous research has provided epigenome-wide insight in regulation of VEGF-A.

METHODS

VEGF-A concentrations of healthy participants from the STANISLAS Family Study (n = 201) were comprehensively assessed for association with DNA methylation. Genome-wide DNA methylation profiles were determined in whole blood DNA using the 450K Infinium BeadChip Array (Illumina). VEGF-A concentration in PBMC extracts was detected using a high-sensitivity multiplex Cytokine Array (Randox Laboratories, UK).

RESULTS

Epigenome-wide association analysis identified 41 methylation sites significantly associated with VEGF-A concentrations derived from PBMC extracts. Twenty CpG sites within 13 chromosomes reached Holm-Bonferroni significance. Significant values ranged from P = 1.08 × 10-7 to P = 5.64 × 10-15.

CONCLUSION

This study exposed twenty significant CpG sites linking DNA methylation to VEGF-A concentration. Methylation detected in promoter regions, such as TPX2 and HAS-1, could explain previously reported associations with the VEGFA gene. Methylation may also help in the understanding of the regulatory mechanisms of other genes located in the vicinity of detected CpG sites.

An Entropy-Based Method for Identifying Mutual Exclusive Driver Genes in Cancer

Cancer in essence is a complex genomic alteration disease which is caused by the somatic mutations during the lifetime. According to previous researches, the first step to overcome cancer is to identify driver genes which can promote carcinogenesis. However, it is still a big challenge to precisely and efficiently extract the cancer related driver genes because the nature of cancer is heterogeneous and there exists tremendously irrelevant passenger mutations which have no function impact on the cancer's development. In this work, we proposed a novel entropy-based method namely EntroRank to identify driver genes by integrating the subcellular localization information and mutual exclusive of variation frequency into the network. EntroRank can take into full consideration different properties of driver genes. Considering the modularity of driver genes, the mutated genes in the network were first clustered into different subgroups according to their located compartments. After that, the structural entropy of the gene in the subgroup was employed to measure its indispensability. Considering mutual exclusive property between driver genes in the modules, relative entropy was utilized to measure the degree of mutual exclusive between two mutated genes in terms of their variation frequency. We applied our method to three different cancers including lung, prostate, and breast cancer. The results show our method not only detect the well-known important drivers but also prioritiz the rare unknown driver genes. Besides, EntroRank can identify driver genes having mutual exclusive property. Compared with other existing methods, our method achieves a better performance for most of cancer types in terms of Precision, Recall, and Fscore.

Identification of potential crucial genes associated with the pathogenesis and prognosis of prostate cancer

Aim: Prostate cancer (PCa) is the sixth leading cause of cancer-related deaths in men throughout the world. This study aimed to investigate genes associated with the pathogenesis and prognosis of PCa. Materials & methods: Data of PCa cases were obtained from public datasets and were analyzed using an integrated bioinformatics strategy. Results: A total of 969 differential expression genes were identified. Moreover, GSE16560 and The Cancer Genome Atlas (TCGA) data showed a prognostic prompt function of the nine-gene signature, as well as in PCa with Gleason 7. Finally, majority of the nine hub genes were associated with drug sensitivity, mutational landscape, immune infiltrates and clinical characteristics of PCa. Conclusion: The nine-gene signature was correlated with drug sensitivity, mutational landscape, immune infiltrates, clinical characteristics and survival from PCa.

ZGRF1 Is Associated with Poor Prognosis in Triple-Negative Breast Cancer and Promotes Cancer Stemness Based on Bioinformatics

Background

Recently, long non-coding RNAs (lncRNAs) are important populations of non-coding RNAs with defined key roles in normal breast development as well as breast tumorigenesis. Triple-negative breast cancer (TNBC) is a particular breast cancer subtype with poor prognosis because of highly invasive and no specific drug treatment yet. Breast cancer stems cells (BCSCs) cause a high risk of invasion, metastasis and drug resistance for breast cancer patients.

Methods

Two microarrays of BCSCs and no-BCSCs were isolated from mammosphere-3D-cultured MCF-7 cells, differentially expressed lncRNAs (DELs) were screened out by Gene Expression Omnibus (GEO). Gene ontology enrichment analysis and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were also performed to analyze DELs features. Using the STRING database to analyze DELs interaction network module to further screen the hub lncRNAs related to tumor stemness and make functional annotations. The expressions of hub DELs were validated using data from The Cancer Genome Atlas database. In addition, the expression analysis and survival analysis were conducted using GEO was utilized to analyze DELs in TNBC using GEPIA database.

Results

A total of 143 aberrantly expressed lncRNAs in BCSCs were identified, and 25 lncRNAs were downregulated and 118 lncRNAs were upregulated compared to non-BCSCs. Up- and downregulated top 3 lncRNAs were selected and verified by RT-PCR. Notably, GO enrichment analysis and KEGG pathway analysis indicated that RNA transport, spliceosome, oxidative phosphorylation, NOD-like receptor signaling pathway, PI3K-Akt signaling pathway, and metabolic pathways may serve important roles in BCSCs. Additionally, the function loss assay indicated that ZGRF1 positively upregulated phenotype and biological functions of BCSCs in vitro. Collectively, our work establishes the lncRNAs signature in BCSCs and these findings assess us with evidence to explore further functionalities of lncRNAs in BCSCs and provide a novel therapeutic strategy for breast cancer.

Conclusion

Our work establishes the lncRNAs signature in BCSCs and these findings assess us with evidence to explore further functionalities of lncRNAs in BCSCs and provide a novel therapeutic strategy for breast cancer. ZGRF1 expression is upregulated in TNBC patients and has a poor prognosis, which can be promising biomarkers.

Machine learning-based analysis of multi-omics data on the cloud for investigating gene regulations

Gene expressions are subtly regulated by quantifiable measures of genetic molecules such as interaction with other genes, methylation, mutations, transcription factor and histone modifications. Integrative analysis of multi-omics data can help scientists understand the condition or patient-specific gene regulation mechanisms. However, analysis of multi-omics data is challenging since it requires not only the analysis of multiple omics data sets but also mining complex relations among different genetic molecules by using state-of-the-art machine learning methods. In addition, analysis of multi-omics data needs quite large computing infrastructure. Moreover, interpretation of the analysis results requires collaboration among many scientists, often requiring reperforming analysis from different perspectives. Many of the aforementioned technical issues can be nicely handled when machine learning tools are deployed on the cloud. In this survey article, we first survey machine learning methods that can be used for gene regulation study, and we categorize them according to five different goals: gene regulatory subnetwork discovery, disease subtype analysis, survival analysis, clinical prediction and visualization. We also summarize the methods in terms of multi-omics input types. Then, we explain why the cloud is potentially a good solution for the analysis of multi-omics data, followed by a survey of two state-of-the-art cloud systems, Galaxy and BioVLAB. Finally, we discuss important issues when the cloud is used for the analysis of multi-omics data for the gene regulation study.

Immune Cell Types and Secreted Factors Contributing to Inflammation-to-Cancer Transition and Immune Therapy Response

Although chronic inflammation increases many cancers' risk, how inflammation facilitates cancer development is still not well studied. Recognizing whether and when inflamed tissues transition to cancerous tissues is of utmost importance. To unbiasedly infer molecular events, immune cell types, and secreted factors contributing to the inflammation-to-cancer (I2C) transition, we develop a computational package called "SwitchDetector" based on liver, gastric, and colon cancer I2C data. Using it, we identify angiogenesis associated with a common critical transition stage for multiple I2C events. Furthermore, we infer infiltrated immune cell type composition and their secreted or suppressed extracellular proteins to predict expression of important transition stage genes. This identifies extracellular proteins that may serve as early-detection biomarkers for pre-cancer and early-cancer stages. They alone or together with I2C hallmark angiogenesis genes are significantly related to cancer prognosis and can predict immune therapy response. The SwitchDetector and I2C database are publicly available at www.inflammation2cancer.org.

Pathway perturbations in signaling networks: Linking genotype to phenotype

Genes and gene products interact with each other to form signal transduction networks in the cell. The interactome networks are under intricate regulation in physiological conditions, but could go awry upon genome instability caused by genetic mutations. In the past decade with next-generation sequencing technologies, an increasing number of genomic mutations have been identified in a variety of disease patients and healthy individuals. As functional and systematic studies on these mutations leap forward, they begin to reveal insights into cellular homeostasis and disease mechanisms. In this review, we discuss recent advances in the field of network biology and signaling pathway perturbations upon genomic changes, and highlight the success of various omics datasets in unraveling genotype-to-phenotype relationships.

PRODIGY: personalized prioritization of driver genes

MOTIVATION

Evolution of cancer is driven by few somatic mutations that disrupt cellular processes, causing abnormal proliferation and tumor development, whereas most somatic mutations have no impact on progression. Distinguishing those mutated genes that drive tumorigenesis in a patient is a primary goal in cancer therapy: Knowledge of these genes and the pathways on which they operate can illuminate disease mechanisms and indicate potential therapies and drug targets. Current research focuses mainly on cohort-level driver gene identification but patient-specific driver gene identification remains a challenge.

METHODS

We developed a new algorithm for patient-specific ranking of driver genes. The algorithm, called PRODIGY, analyzes the expression and mutation profiles of the patient along with data on known pathways and protein-protein interactions. Prodigy quantifies the impact of each mutated gene on every deregulated pathway using the prize-collecting Steiner tree model. Mutated genes are ranked by their aggregated impact on all deregulated pathways.

RESULTS

In testing on five TCGA cancer cohorts spanning >2500 patients and comparison to validated driver genes, Prodigy outperformed extant methods and ranking based on network centrality measures. Our results pinpoint the pleiotropic effect of driver genes and show that Prodigy is capable of identifying even very rare drivers. Hence, Prodigy takes a step further toward personalized medicine and treatment.

AVAILABILITY AND IMPLEMENTATION

The Prodigy R package is available at: https://github.com/Shamir-Lab/PRODIGY.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Biclustering-based association rule mining approach for predicting cancer-associated protein interactions

Protein-protein interactions (PPIs) have been widely used to understand different biological processes and cellular functions associated with several diseases like cancer. Although some cancer-related protein interaction databases are available, lack of experimental data and conflicting PPI data among different available databases have slowed down the cancer research. Therefore, in this study, the authors have focused on various proteins that are directly related to different types of cancer disease. They have prepared a PPI database between cancer-associated proteins with the rest of the human proteins. They have also incorporated the annotation type and direction of each interaction. Subsequently, a biclustering-based association rule mining algorithm is applied to predict new interactions with type and direction. This study shows the prediction power of association rule mining algorithm over the traditional classifier model without choosing a negative data set. The time complexity of the biclustering-based association rule mining is also analysed and compared to traditional association rule mining. The authors are able to discover 38 new PPIs which are not present in the cancer database. The biological relevance of these newly predicted interactions is analysed by published literature. Recognition of such interactions may accelerate a way of developing new drugs to prevent different cancer-related diseases.

Similarity of therapeutic networks induced by a multi-component herbal remedy, Ukgansan, in neurovascular unit cells

The neurovascular unit, which includes neurons, glial cells, and vascular cells, plays crucial roles in the onset and progression of Alzheimer’s disease (AD). Therefore, effective drugs against AD should be able to target the multi-cellular neurovascular unit and the therapeutic relationships among neurovascular cells should be defined. Here, we examined the therapeutic effects of Ukgansan (UGS), an herbal remedy with multi-targeting capabilities, using in vitro neurovascular unit models and an in vivo model of AD. In addition, we compared the therapeutic networks induced by UGS and its components in different neurovascular cell types. We found that UGS and its components protected neurovascular cells against diverse damaging agents and improved the behavioral patterns of AD model mice. A comparison of UGS- or its components-induced therapeutic networks, constructed from high-throughput data on gene expression, pathway activity, and protein phosphorylation, revealed similarities among neurovascular cell types, especially between BV-2 microglia and HBVP (human brain vascular pericytes). These findings, together with the functional connections between neurovascular cells, can explain the therapeutic effects of UGS. Furthermore, they suggest underlying similarities in the therapeutic mechanisms in different neurovascular cell types.

Host factor prioritization for pan-viral genetic perturbation screens using random intercept models and network propagation

Genetic perturbation screens using RNA interference (RNAi) have been conducted successfully to identify host factors that are essential for the life cycle of bacteria or viruses. So far, most published studies identified host factors primarily for single pathogens. Furthermore, often only a small subset of genes, e.g., genes encoding kinases, have been targeted. Identification of host factors on a pan-pathogen level, i.e., genes that are crucial for the replication of a diverse group of pathogens has received relatively little attention, despite the fact that such common host factors would be highly relevant, for instance, for devising broad-spectrum anti-pathogenic drugs. Here, we present a novel two-stage procedure for the identification of host factors involved in the replication of different viruses using a combination of random effects models and Markov random walks on a functional interaction network. We first infer candidate genes by jointly analyzing multiple perturbations screens while at the same time adjusting for high variance inherent in these screens. Subsequently the inferred estimates are spread across a network of functional interactions thereby allowing for the analysis of missing genes in the biological studies, smoothing the effect sizes of previously found host factors, and considering a priori pathway information defined over edges of the network. We applied the procedure to RNAi screening data of four different positive-sense single-stranded RNA viruses, Hepatitis C virus, Chikungunya virus, Dengue virus and Severe acute respiratory syndrome coronavirus, and detected novel host factors, including UBC, PLCG1, and DYRK1B, which are predicted to significantly impact the replication cycles of these viruses. We validated the detected host factors experimentally using pharmacological inhibition and an additional siRNA screen and found that some of the predicted host factors indeed influence the replication of these pathogens.

Network Embedding the Protein-Protein Interaction Network for Human Essential Genes Identification

Essential genes are a group of genes that are indispensable for cell survival and cell fertility. Studying human essential genes helps scientists reveal the underlying biological mechanisms of a human cell but also guides disease treatment. Recently, the publication of human essential gene data makes it possible for researchers to train a machine-learning classifier by using some features of the known human essential genes and to use the classifier to predict new human essential genes. Previous studies have found that the essentiality of genes closely relates to their properties in the protein–protein interaction (PPI) network. In this work, we propose a novel supervised method to predict human essential genes by network embedding the PPI network. Our approach implements a bias random walk on the network to get the node network context. Then, the node pairs are input into an artificial neural network to learn their representation vectors that maximally preserves network structure and the properties of the nodes in the network. Finally, the features are put into an SVM classifier to predict human essential genes. The prediction results on two human PPI networks show that our method achieves better performance than those that refer to either genes’ sequence information or genes’ centrality properties in the network as input features. Moreover, it also outperforms the methods that represent the PPI network by other previous approaches.

netNMF-sc: leveraging gene-gene interactions for imputation and dimensionality reduction in single-cell expression analysis

Single-cell RNA-sequencing (scRNA-seq) enables high-throughput measurement of RNA expression in single cells. However, because of technical limitations, scRNA-seq data often contain zero counts for many transcripts in individual cells. These zero counts, or dropout events, complicate the analysis of scRNA-seq data using standard methods developed for bulk RNA-seq data. Current scRNA-seq analysis methods typically overcome dropout by combining information across cells in a lower-dimensional space, leveraging the observation that cells generally occupy a small number of RNA expression states. We introduce netNMF-sc, an algorithm for scRNA-seq analysis that leverages information across both cells and genes. netNMF-sc learns a low-dimensional representation of scRNA-seq transcript counts using network-regularized non-negative matrix factorization. The network regularization takes advantage of prior knowledge of gene–gene interactions, encouraging pairs of genes with known interactions to be nearby each other in the low-dimensional representation. The resulting matrix factorization imputes gene abundance for both zero and nonzero counts and can be used to cluster cells into meaningful subpopulations. We show that netNMF-sc outperforms existing methods at clustering cells and estimating gene–gene covariance using both simulated and real scRNA-seq data, with increasing advantages at higher dropout rates (e.g., >60%). We also show that the results from netNMF-sc are robust to variation in the input network, with more representative networks leading to greater performance gains.

Maternal environmental exposure to bisphenols and epigenome-wide DNA methylation in infant cord blood

Maternal prenatal exposures, including bisphenol A (BPA), are associated with offspring's risk of disease later in life. Alterations in DNA methylation may be a mechanism through which altered prenatal conditions (e.g. maternal exposure to environmental toxicants) elicit this disease risk. In the Michigan Mother and Infant Pairs Cohort, maternal first-trimester urinary BPA, bisphenol F, and bisphenol S concentrations were tested for association with DNA methylation patterns in infant umbilical cord blood leukocytes (N = 69). We used the Illumina Infinium MethylationEPIC BeadChip to quantitatively evaluate DNA methylation across the epigenome; 822 020 probes passed pre-processing and quality checks. Single-site DNA methylation and bisphenol models were adjusted for infant sex, estimated cell-type proportions (determined using cell-type estimation algorithm), and batch as covariates. Thirty-eight CpG sites [false discovery rate (FDR) <0.05] were significantly associated with maternal BPA exposure. Increasing BPA concentrations were associated with lower DNA methylation at 87% of significant sites. BPA exposure associated DNA methylation sites were enriched for 38 pathways significant at FDR <0.05. The pathway or gene-set with the greatest odds of enrichment for differential methylation (FDR <0.05) was type I interferon receptor binding. This study provides a novel understanding of fetal response to maternal bisphenol exposure through epigenetic change.

Identifying driver genes involving gene dysregulated expression, tissue-specific expression and gene-gene network

BACKGROUND

Cancer as a kind of genomic alteration disease each year deprives many people's life. The biggest challenge to overcome cancer is to identify driver genes that promote the cancer development from a huge amount of passenger mutations that have no effect on the selective growth advantage of cancer. In order to solve those problems, some researchers have started to focus on identification of driver genes by integrating networks with other biological information. However, more efforts should be needed to improve the prediction performance.

METHODS

Considering the facts that driver genes have impact on expression of their downstream genes, they likely interact with each other to form functional modules and those modules should tend to be expressed similarly in the same tissue. We proposed a novel model named by DyTidriver to identify driver genes through involving the gene dysregulated expression, tissue-specific expression and variation frequency into the human functional interaction network (e.g. human FIN).

RESULTS

This method was applied on 974 breast, 316 prostate and 230 lung cancer patients. The consequence shows our method outperformed other five existing methods in terms of Fscore, Precision and Recall values. The enrichment and cociter analysis illustrate DyTidriver can not only identifies the driver genes enriched in some significant pathways but also has the capability to figure out some unknown driver genes.

CONCLUSION

The final results imply that driver genes are those that impact more dysregulated genes and express similarly in the same tissue.

KRCC1: A potential therapeutic target in ovarian cancer

Using a systems biology approach to prioritize potential points of intervention in ovarian cancer, we identified the lysine rich coiled-coil 1 (KRCC1), as a potential target. High-grade serous ovarian cancer patient tumors and cells express significantly higher levels of KRCC1 which correlates with poor overall survival and chemoresistance. We demonstrate that KRCC1 is predominantly present in the chromatin-bound nuclear fraction, interacts with HDAC1, HDAC2, and with the serine-threonine phosphatase PP1CC. Silencing KRCC1 inhibits cellular plasticity, invasive properties, and potentiates apoptosis resulting in reduced tumor growth. These phenotypes are associated with increased acetylation of histones and with increased phosphorylation of H2AX and CHK1, suggesting the modulation of transcription and DNA damage that may be mediated by the action of HDAC and PP1CC, respectively. Hence, we address an urgent need to develop new targets in cancer.

CBNA: A control theory based method for identifying coding and non-coding cancer drivers

A key task in cancer genomics research is to identify cancer driver genes. As these genes initialise and progress cancer, understanding them is critical in designing effective cancer interventions. Although there are several methods developed to discover cancer drivers, most of them only identify coding drivers. However, non-coding RNAs can regulate driver mutations to develop cancer. Hence, novel methods are required to reveal both coding and non-coding cancer drivers. In this paper, we develop a novel framework named Controllability based Biological Network Analysis (CBNA) to uncover coding and non-coding cancer drivers (i.e. miRNA cancer drivers). CBNA integrates different genomic data types, including gene expression, gene network, mutation data, and contains a two-stage process: (1) Building a network for a condition (e.g. cancer condition) and (2) Identifying drivers. The application of CBNA to the BRCA dataset demonstrates that it is more effective than the existing methods in detecting coding cancer drivers. In addition, CBNA also predicts 17 miRNA drivers for breast cancer. Some of these predicted miRNA drivers have been validated by literature and the rest can be good candidates for wet-lab validation. We further use CBNA to detect subtype-specific cancer drivers and several predicted drivers have been confirmed to be related to breast cancer subtypes. Another application of CBNA is to discover epithelial-mesenchymal transition (EMT) drivers. Of the predicted EMT drivers, 7 coding and 6 miRNA drivers are in the known EMT gene lists.

Cancer is a disease of cells in human body and it causes a high rate of deaths worldwide. There has been evidence that coding and non-coding RNAs are key players in the initialisation and progression of cancer. These coding and non-coding RNAs are considered as cancer drivers. To design better diagnostic and therapeutic plans for cancer patients, we need to know the roles of cancer drivers in cancer development as well as their regulatory mechanisms in the human body. In this study, we propose a novel framework to identify coding and non-coding cancer drivers (i.e. miRNA cancer drivers). The proposed framework is applied to the breast cancer dataset for identifying drivers of breast cancer. Comparing our method with existing methods in predicting coding cancer drivers, our method shows a better performance. Several miRNA cancer drivers predicted by our method have already been validated by literature. The predicted cancer drivers by our method could be a potential source for further wet-lab experiments to discover the causes of cancer. In addition, the proposed method can be used to detect drivers of cancer subtypes and drivers of the epithelial-mesenchymal transition in cancer.

A novel network control model for identifying personalized driver genes in cancer

Although existing computational models have identified many common driver genes, it remains challenging to identify the personalized driver genes by using samples of an individual patient. Recently, the methods of exploiting the structure-based control principles of complex networks provide new clues for identifying minimum number of driver nodes to drive the state transition of large-scale complex networks from an initial state to the desired state. However, the structure-based network control methods cannot be directly applied to identify the personalized driver genes due to the unknown network dynamics of the personalized system. Here we proposed the personalized network control model (PNC) to identify the personalized driver genes by employing the structure-based network control principle on genetic data of individual patients. In PNC model, we firstly presented a paired single sample network construction method to construct the personalized state transition network for capturing the phenotype transitions between healthy and disease states. Then, we designed a novel structure-based network control method from the Feedback Vertex Sets-based control perspective to identify the personalized driver genes. The wide experimental results on 13 cancer datasets from The Cancer Genome Atlas firstly showed that PNC model outperforms current state-of-the-art methods, in terms of F-measures for identifying cancer driver genes enriched in the gold-standard cancer driver gene lists. Furthermore, these results showed that personalized driver genes can be explored by their network characteristics even when they are hidden factors in transcription and mutation profiles. Our PNC gives novel insights and useful tools into understanding the tumor heterogeneity in cancer. The PNC package and data resources used in this work can be freely downloaded from https://github.com/NWPU-903PR/PNC.

Notably there may be unique personalized driver genes for an individual patient in cancer. Identifying personalized driver genes that lead to particular cancer initiation and progression of individual patient is one of the biggest challenges in precision medicine. However, most methods for cancer driver genes identification have focused mainly on the cohort information rather than on individual information and fail to identify personalized driver genes. We here proposed personalized network control model (PNC) to identify personalized driver genes by applying the structure based network control principle on personalized data of individual patients. By considering the progression from the healthy state to the disease state as the network control problem, our PNC aims to detect a small number of personalized driver genes that are altered in response to input signals for triggering the state transition in individual patients on expression level. The impetus behind PNC contains two main respects. One is to design a paired single sample network construction method (namely Paired-SSN) for constructing personalized state transition networks to capture the phenotypic transitions between normal and disease attractors. The other one is to develop a novel structure based network control method (namely NCUA) on personalized state transition networks for identifying personalized driver genes which can drive individual patient system state from healthy state to disease state through oncogene activations. Each part of the proposed method has been deeply examined to be efficient. Compared with other existing models, our PNC shows a higher performance in terms of F-measures of the cancer driver genes in the well-known Cancer Census Genes (CCG) and Network of Cancer Genes (NCG). The wide experimental results on multiple cancer datasets highlight that sample specific network theory and structure based network control theory can contribute to identifying personalized driver genes in cancer.

Identification of the key differentially expressed genes and pathways involved in neutrophilia

Polymorphonuclear neutrophils (PMNs) are the most important determinants in the acute inflammatory response. Pathologically increased numbers of PMNs in the circulation or specific tissues (or both) lead to neutrophilia. However, the genes expressed and pathways involved in neutrophilia have yet to be elucidated. By analysis of three public microarray datasets related to neutrophilia (GSE64457, GSE54644, and GSE94923) and evaluation by gene ontology, pathway enrichment, protein-protein interaction networks, and hub genes analysis using multiple methods (DAVID, PATHER, Reactome, STRING, Reactome FI Plugin, and CytoHubba in Cytoscape), we identified the commonly up-regulated and down-regulated different expressed genes. We also discovered that multiple signaling pathways (IL-mediated, LPS-mediated, TNF-α, TLR cascades, MAPK, and PI3K-Akt) were involved in PMN regulation. Our findings suggest that the commonly expressed genes involved in regulation of multiple pathways were the underlying molecular mechanisms in the development of inflammatory, autoimmune, and hematologic diseases that share the common phenotypic characteristics of increased numbers of PMNs. Taken together, these data suggest that these genes are involved in the regulation of neutrophilia and that the corresponding gene products could serve as potential biomarkers and/or therapeutic targets for neutrophilia.

SigHotSpotter: scRNA-seq-based computational tool to control cell subpopulation phenotypes for cellular rejuvenation strategies

SUMMARY

Single-cell RNA-sequencing is increasingly employed to characterize disease or ageing cell subpopulation phenotypes. Despite exponential increase in data generation, systematic identification of key regulatory factors for controlling cellular phenotype to enable cell rejuvenation in disease or ageing remains a challenge. Here, we present SigHotSpotter, a computational tool to predict hotspots of signaling pathways responsible for the stable maintenance of cell subpopulation phenotypes, by integrating signaling and transcriptional networks. Targeted perturbation of these signaling hotspots can enable precise control of cell subpopulation phenotypes. SigHotSpotter correctly predicts the signaling hotspots with known experimental validations in different cellular systems. The tool is simple, user-friendly and is available as web-server or as stand-alone software. We believe SigHotSpotter will serve as a general purpose tool for the systematic prediction of signaling hotspots based on single-cell RNA-seq data, and potentiate novel cell rejuvenation strategies in the context of disease and ageing.

AVAILABILITY AND IMPLEMENTATION

SigHotSpotter is at https://SigHotSpotter.lcsb.uni.lu as a web tool. Source code, example datasets and other information are available at https://gitlab.com/srikanth.ravichandran/sighotspotter.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Survival-associated alternative splicing signatures in esophageal carcinoma

Alternative splicing (AS), a major mechanism for the enhancement of transcriptome and proteome diversity, has been widely demonstrated to be involved in the full spectrum of oncogenic processes. High-throughput sequencing technology and the rapid accumulation of clinical data sets have provided an opportunity to systemically analyze the association between messenger RNA AS variants and patient clinical outcomes. Here, we compared differentially spliced AS transcripts between esophageal carcinoma (ESCA) and non-tumor tissues, profiled genome-wide survival-associated AS events in 87 patients with esophageal adenocarcinoma (EAC) and 79 patients with esophageal squamous cell carcinoma (ESCC) using The Cancer Genome Atlas (TCGA) RNA-seq data set, and constructed predictive models as well as splicing regulation networks by integrated bioinformatic analysis. A total of 2326 AS events in 1738 genes and 1812 AS events in 1360 genes were determined to be significantly associated with overall survival (OS) of patients in the EAC and ESCC cohorts, respectively, including some essential participants in the oncogenic process. The predictive model of each splice type performed reasonably well in distinguishing good and poor outcomes of patients with esophageal cancer, and values for the area under curve reached 0.942 and 0.815 in the EAC exon skip predictive model and the ESCC alternate acceptor site predictive model, respectively. The splicing regulation networks revealed an interesting correlation between survival-associated splicing factors and prognostic AS genes. In summary, we created prognostic models for patients with esophageal cancer based on AS signatures and constructed novel splicing correlation networks.

Similarities and Differences of Blood N-Glycoproteins in Five Solid Carcinomas at Localized Clinical Stage Analyzed by SWATH-MS

Cancer is mostly incurable when diagnosed at a metastatic stage, making its early detection via blood proteins of immense clinical interest. Proteomic changes in tumor tissue may lead to changes detectable in the protein composition of circulating blood plasma. Using a proteomic workflow combining N-glycosite enrichment and SWATH mass spectrometry, we generate a data resource of 284 blood samples derived from patients with different types of localized-stage carcinomas and from matched controls. We observe whether the changes in the patient's plasma are specific to a particular carcinoma or represent a generic signature of proteins modified uniformly in a common, systemic response to many cancers. A quantitative comparison of the resulting N-glycosite profiles discovers that proteins related to blood platelets are common to several cancers (e.g., THBS1), whereas others are highly cancer-type specific. Available proteomics data, including a SWATH library to study N-glycoproteins, will facilitate follow-up biomarker research into early cancer detection.

A Computational Framework for Predicting Direct Contacts and Substructures within Protein Complexes

Understanding the physical arrangement of subunits within protein complexes potentially provides valuable clues about how the subunits work together and how the complexes function. The majority of recent research focuses on identifying protein complexes as a whole and seldom studies the inner structures within complexes. In this study, we propose a computational framework to predict direct contacts and substructures within protein complexes. In this framework, we first train a supervised learning model of l₂-regularized logistic regression to learn the patterns of direct and indirect interactions within complexes, from where physical subunit interaction networks are predicted. Then, to infer substructures within complexes, we apply a graph clustering method (i.e., maximum modularity clustering (MMC)) and a gene ontology (GO) semantic similarity based functional clustering on partially- and fully-connected networks, respectively. Computational results show that the proposed framework achieves fairly good performance of cross validation and independent test in terms of detecting direct contacts between subunits. Functional analyses further demonstrate the rationality of partitioning the subunits into substructures via the MMC algorithm and functional clustering.

Mapping genetic interactions in cancer: a road to rational combination therapies

The discovery of synthetic lethal interactions between poly (ADP-ribose) polymerase (PARP) inhibitors and BRCA genes, which are involved in homologous recombination, led to the approval of PARP inhibition as a monotherapy for patients with BRCA1/2-mutated breast or ovarian cancer. Studies following the initial observation of synthetic lethality demonstrated that the reach of PARP inhibitors is well beyond just BRCA1/2 mutants. Insights into the mechanisms of action of anticancer drugs are fundamental for the development of targeted monotherapies or rational combination treatments that will synergize to promote cancer cell death and overcome mechanisms of resistance. The development of targeted therapeutic agents is premised on mapping the physical and functional dependencies of mutated genes in cancer. An important part of this effort is the systematic screening of genetic interactions in a variety of cancer types. Until recently, genetic-interaction screens have relied either on the pairwise perturbations of two genes or on the perturbation of genes of interest combined with inhibition by commonly used anticancer drugs. Here, we summarize recent advances in mapping genetic interactions using targeted, genome-wide, and high-throughput genetic screens, and we discuss the therapeutic insights obtained through such screens. We further focus on factors that should be considered in order to develop a robust analysis pipeline. Finally, we discuss the integration of functional interaction data with orthogonal methods and suggest that such approaches will increase the reach of genetic-interaction screens for the development of rational combination therapies.

Acute myeloid leukemia-induced T-cell suppression can be reversed by inhibition of the MAPK pathway

Acute myeloid leukemia (AML) remains difficult to treat due to mutational heterogeneity and the development of resistance to therapy. Targeted agents, such as MEK inhibitors, may be incorporated into treatment; however, the impact of MEK inhibitors on the immune microenvironment in AML is not well understood. A greater understanding of the implications of MEK inhibition on immune responses may lead to a greater understanding of immune evasion and more rational combinations with immunotherapies. This study describes the impact of trametinib on both T cells and AML blast cells by using an immunosuppressive mouse model of AML and primary patient samples. We also used a large AML database of functional drug screens to understand characteristics of trametinib-sensitive samples. In the mouse model, trametinib increased T-cell viability and restored T-cell proliferation. Importantly, we report greater proliferation in the CD8+CD44+ effector subpopulation and impaired activation of CD8+CD62L+ naive cells. Transcriptome analysis revealed that trametinib-sensitive samples have an inflammatory gene expression profile, and we also observed increased programmed cell death ligand 1 (PD-L1) expression on trametinib-sensitive samples. Finally, we found that trametinib consistently reduced PD-L1 and PD-L2 expression in a dose-dependent manner on the myeloid population. Altogether, our data present greater insight into the impact of trametinib on the immune microenvironment and characteristics of trametinib-sensitive patient samples.

Neglog: Homology-Based Negative Data Sampling Method for Genome-Scale Reconstruction of Human Protein-Protein Interaction Networks

Rapid reconstruction of genome-scale protein-protein interaction (PPI) networks is instrumental in understanding the cellular processes and disease pathogenesis and drug reactions. However, lack of experimentally verified negative data (i.e., pairs of proteins that do not interact) is still a major issue that needs to be properly addressed in computational modeling. In this study, we take advantage of the very limited experimentally verified negative data from Negatome to infer more negative data for computational modeling. We assume that the paralogs or orthologs of two non-interacting proteins also do not interact with high probability. We coin an assumption as "Neglog" this assumption is to some extent supported by paralogous/orthologous structure conservation. To reduce the risk of bias toward the negative data from Negatome, we combine Neglog with less biased random sampling according to a certain ratio to construct training data. L₂-regularized logistic regression is used as the base classifier to counteract noise and train on a large dataset. Computational results show that the proposed Neglog method outperforms pure random sampling method with sound biological interpretability. In addition, we find that independent test on negative data is indispensable for bias control, which is usually neglected by existing studies. Lastly, we use the Neglog method to validate the PPIs in STRING, which are supported by gene ontology (GO) enrichment analyses.

Systematic Profiling of Alternative mRNA Splicing Signature for Predicting Glioblastoma Prognosis

Emerging evidence suggests that alternative splicing (AS) is modified in cancer and is associated with cancer progression. Systematic analysis of AS signature in glioblastoma (GBM) is lacking and is greatly needed. We profiled genome-wide AS events in 498 GBM patients in TCGA using RNA-seq data, and splicing network and prognostic predictor were built by integrated bioinformatics analysis. Among 45,610 AS events in 10,434 genes, we detected 1,829 AS events in 1,311 genes, and 1,667 AS events in 1,146 genes that were significantly associated with overall survival and disease-free survival of GBM patients, respectively. Five potential feature genes, S100A4, ECE2, CAST, ASPH, and LY6K, were discovered after network mining as well as correlation analysis between AS and gene expression, most of which were related to carcinogenesis and development. Multivariate survival model analysis indicated that these five feature genes could classify the prognosis at AS event and gene expression level. This report opens up a new avenue for exploration of the pathogenesis of GBM through AS, thus more precisely guiding clinical treatment and prognosis judgment.

Network-based integration of multi-omics data for prioritizing cancer genes

Motivation

Several molecular events are known to be cancer-related, including genomic aberrations, hypermethylation of gene promoter regions and differential expression of microRNAs. These aberration events are very heterogeneous across tumors and it is poorly understood how they affect the molecular makeup of the cell, including the transcriptome and proteome. Protein interaction networks can help decode the functional relationship between aberration events and changes in gene and protein expression.

Results

We developed NetICS (Network-based Integration of Multi-omics Data), a new graph diffusion-based method for prioritizing cancer genes by integrating diverse molecular data types on a directed functional interaction network. NetICS prioritizes genes by their mediator effect, defined as the proximity of the gene to upstream aberration events and to downstream differentially expressed genes and proteins in an interaction network. Genes are prioritized for individual samples separately and integrated using a robust rank aggregation technique. NetICS provides a comprehensive computational framework that can aid in explaining the heterogeneity of aberration events by their functional convergence to common differentially expressed genes and proteins. We demonstrate NetICS’ competitive performance in predicting known cancer genes and in generating robust gene lists using TCGA data from five cancer types.

Availability and implementation

NetICS is available at https://github.com/cbg-ethz/netics.

Supplementary information

Supplementary data are available at Bioinformatics online.

Proteomic profiling of extracellular vesicles allows for human breast cancer subtyping

Extracellular vesicles (EVs) are a potential source of disease-associated biomarkers for diagnosis. In breast cancer, comprehensive analyses of EVs could yield robust and reliable subtype-specific biomarkers that are still critically needed to improve diagnostic routines and clinical outcome. Here, we show that proteome profiles of EVs secreted by different breast cancer cell lines are highly indicative of their respective molecular subtypes, even more so than the proteome changes within the cancer cells. Moreover, we detected molecular evidence for subtype-specific biological processes and molecular pathways, hyperphosphorylated receptors and kinases in connection with the disease, and compiled a set of protein signatures that closely reflect the associated clinical pathophysiology. These unique features revealed in our work, replicated in clinical material, collectively demonstrate the potential of secreted EVs to differentiate between breast cancer subtypes and show the prospect of their use as non-invasive liquid biopsies for diagnosis and management of breast cancer patients.

Analyzing the regulation of miRNAs on protein-protein interaction network in Hodgkin lymphoma

Background

Hodgkin Lymphoma (HL) is a type of aggressive malignancy in lymphoma that has high incidence in young adults and elderly patients. Identification of reliable diagnostic markers and efficient therapeutic targets are especially important for the diagnosis and treatment of HL. Although many HL-related molecules have been identified, our understanding on the molecular mechanisms underlying the disease is still far from complete due to its complex and heterogeneous characteristics. In such situation, exploring the molecular mechanisms underlying HL via systems biology approaches provides a promising option. In this study, we try to elucidate the molecular mechanisms related to the disease and identify potential pharmaceutical targets from a network-based perspective.

Results

We constructed a series of network models. Based on the analysis of these networks, we attempted to identify the biomarkers and elucidate the molecular mechanisms underlying HL. Initially, we built three different but related protein networks, i.e., background network, HL-basic network and HL-specific network. By analyzing these three networks, we investigated the connection characteristic of the HL-related proteins. Subsequently, we explored the miRNA regulation on HL-specific network and analyzed three kinds of simple regulation patterns, i.e., co-regulation of protein pairs, as well as the direct and indirect regulation of triple proteins. Finally, we constructed a simplified protein network combined with the regulation of miRNAs on proteins to better understand the relation between HL-related proteins and miRNAs.

Conclusions

We find that the HL-related proteins are more likely to connect with each other compared to other proteins. Moreover, the HL-specific network can be further divided into five sub-networks and 49 proteins as the backbone of HL-specific network make up and connect these 5 sub-networks. Thus, they may be closely associated with HL. In addition, we find that the co-regulation of protein pairs is the main regulatory pattern of miRNAs on the protein network in the HL-specific network. According to the regulation of miRNA on protein network, we have identified 5 core miRNAs as the potential biomarkers for diagnostic of HL. Finally, several protein pathways have been identified to closely associated with HL, which provides deep insights into underlying mechanism of HL.

Electronic supplementary material

The online version of this article (10.1186/s12859-019-3041-9) contains supplementary material, which is available to authorized users.

Systematic analysis of survival-associated alternative splicing signatures uncovers prognostic predictors for head and neck cancer

BACKGROUND

Previous studies have shown that alternative splicing (AS) plays a key role in carcinogenesis and prognosis of cancer. However, systematic profiles of AS signatures in head and neck cancer (HNC) have not yet been reported.

METHODS

In this study, AS data, RNA-Seq data, and corresponding clinicopathological information of 489 HNC patients were downloaded from The Cancer Genome Atlas. Univariate and multivariate Cox regression analyses were performed to screen for survival-associated AS events. Functional and pathway enrichment analysis was also performed. The prognostic models and splicing networks were constructed using integrated bioinformatics analysis tools.

RESULTS

Among the 42,849 alternating splicing events identified in 10,121 genes, 5,165 survival-associated AS events in 2,419 genes were observed in univariate Cox regression analysis. Among the seven types, alternate terminator events were the most powerful prognostic factors. Multivariate Cox analysis was then used to screen for the AS genes with prognostic value. Four candidate genes (TPM1, CLASRP, PRRC1, and DNASE1L1) were found to be independent prognostic factors for HNC patients. A prognostic prediction model was built based on the four genes. The area under the receiver operating characteristic risk score curve for predicting the survival status of HNC patients was 0.704. In addition, splicing interaction network indicated that the splicing factors have significant functions in HNC.

CONCLUSION

A comprehensive analysis of AS events in HNC was performed. A powerful prognostic predictor for HNC patients was established based on AS events could.