Dysfunctions associated with methylation, microRNA expression and gene expression in lung cancer

Bayesian tensor factorization-drive breast cancer subtyping by integrating multi-omics data

Breast cancer is a highly heterogeneous disease. Subtyping the disease and identifying the genomic features driving these subtypes are critical for precision oncology for breast cancer. This study focuses on developing a new computational approach for breast cancer subtyping. We proposed to use Bayesian tensor factorization (BTF) to integrate multi-omics data of breast cancer, which include expression profiles of RNA-sequencing, copy number variation, and DNA methylation measured on 762 breast cancer patients from The Cancer Genome Atlas. We applied a consensus clustering approach to identify breast cancer subtypes using the factorized latent features by BTF. Subtype-specific survival patterns of the breast cancer patients were evaluated using Kaplan-Meier (KM) estimators. The proposed approach was compared with other state-of-the-art approaches for cancer subtyping. The BTF-subtyping analysis identified 17 optimized latent components, which were used to reveal six major breast cancer subtypes. Out of all different approaches, only the proposed approach showed distinct survival patterns (p < 0.05). Statistical tests also showed that the identified clusters have statistically significant distributions. Our results showed that the proposed approach is a promising strategy to efficiently use publicly available multi-omics data to identify breast cancer subtypes.

A novel method to identify gene interaction patterns

BACKGROUND

Gene interaction patterns, including modules and motifs, can be used to identify cancer specific biomarkers and to reveal the mechanism of tumorigenesis. Most of the existing module network inferencing methods focus on gene independent functional patterns, while the studies of overlapping characteristics between modules are lacking. The objective of this study was to reveal the functional overlapping patterns in gene modules, helping elucidate the regulatory relationship between overlapping genes and communities, as well as to explore cancer formation and progression.

RESULTS

We analyzed six cancer datasets from The Cancer Genome Atlas and obtained three kinds of gene functional modules for each cancer, including Independent-Community, Dependent-Community and Merged-Community. In the six cancers, 59(3.5%) Independent-Communities were identified, while 1631(96.5%) Dependent-Communities were acquired. Compared with Lemon-Tree and K-Means, the gene communities identified by our method were enriched in more known GO categories with lower p-values. Meanwhile, those identified distinguishing communities can significantly distinguish the survival prognostic of patients by Kaplan-Meier analysis. Furthermore, identified driver genes in the gene communities can be considered as biomarkers which can accurately distinguish the tumour or normal samples for each cancer type.

CONCLUSIONS

In all identified communities, Dependent-Communities are the majority. Our method is more effective than the other two methods which do not consider the overlapping characteristics of modules. This indicates that overlapping genes are located in different specific functional groups, and a communication bridge is established between the communities to construct a comprehensive carcinogenesis.

The Serum MicroRNA Signatures for Pancreatic Cancer Detection and Operability Evaluation

Pancreatic cancer (PC) has high morbidity and mortality. It is the fourth leading cause of cancer death. Its diagnosis and treatment are difficult. Liquid biopsy makes early diagnosis of pancreatic cancer possible. We analyzed the expression profiles of 2,555 serum miRNAs in 100 pancreatic cancer patients and 150 healthy controls. With advanced feature selection methods, we identified 13 pancreatic cancer signature miRNAs that can classify the pancreatic cancer patients and healthy controls. For pancreatic cancer treatment, operation is still the first choice. But many pancreatic cancer patients are already inoperable. Therefore, we compared the 79 inoperable and 21 operable patients and identified 432 miRNAs that can predict whether a pancreatic cancer patient was operable. The functional analysis of the 13 pancreatic cancer signatures and the 432 operability miRNAs revealed the molecular mechanisms of pancreatic cancer and shield light on the diagnosis and therapy of pancreatic cancer in clinical practice.

Integrative Analysis of Methylation and Gene Expression in Lung Adenocarcinoma and Squamous Cell Lung Carcinoma

Lung cancer is a highly prevalent type of cancer with a poor 5-year survival rate of about 4-17%. Eighty percent lung cancer belongs to non-small-cell lung cancer (NSCLC). For a long time, the treatment of NSCLC has been mostly guided by tumor stage, and there has been no significant difference between the therapy strategy of lung adenocarcinoma (LUAD) and squamous cell lung carcinoma (SCLC), the two major subtypes of NSCLC. In recent years, important molecular differences between LUAD and SCLC are increasingly identified, indicating that targeted therapy will be more and more histologically specific in the future. To investigate the LUAD and SCLC difference on multi-omics scale, we analyzed the methylation and gene expression data together. With the Boruta method to remove irrelevant features and the MCFS (Monte Carlo Feature Selection) method to identify the significantly important features, we identified 113 key methylation features and 23 key gene expression features. HNF1B and TP63 were found to be dysfunctional on both methylation and gene expression levels. The experimentally determined interaction network suggested that TP63 may play an important role in connecting methylation genes and expression genes. Many of the discovered signature genes have been supported by literature. Our results may provide directions of precision diagnosis and therapy of LUAD and SCLC.

Smoking alters the evolutionary trajectory of non-small cell lung cancer

Smoking is the biggest risk factor for lung cancer. Smokers have a much higher chance of developing lung tumors with a worse survival rate; however, non-smokers also develop lung tumors. A number of questions remain including the underlying difference between smoker and non-smoker lung cancer patients and the involvement of genetic and epigenetic processes in tumor development. The present study analyzed the mutation data of 100 non-small cell lung cancer (NSCLC) patients, 12 non-smokers, 48 ex-smokers and 40 smokers, from Tracking Non-Small Cell Lung Cancer Evolution through Therapy Consortium. A total of 68 genes exhibited different mutation patterns across non-smokers, ex-smokers and smokers. A number of these 68 genes encode membrane proteins with biological regulation, metabolic process, and response to stimulus functions. For each group of patients, the top 10 most frequently mutated genes were selected and their oncogenetic tree inferred, which reflected how the genes evolve during tumor genesis. By comparing the oncogenetic trees of non-smokers and smokers, it was identified that in non-smokers, the mutation of epidermal growth factor receptor (EGFR) was an early genetic alteration event and EGFR was the key driver, but in smokers, the mutation of titin (TTN) was more important. Based on network analysis, TTN can interact with spectrin α erythrocytic 1 through calmodulin 2 and troponin C1. These genetic differences during tumorigenesis of non-smoker and smoker lung cancer patients provided novel insights into the effects of smoking on the evolutionary trajectory of non-small cell lung cancer and may prove helpful for targeted therapy of different lung cancer subtypes.

miR-137 inhibits the proliferation of human non-small cell lung cancer cells by targeting SRC3

Non-small cell lung cancer (NSCLC) is the most common type of lung cancer. The results of the present study demonstrate that high expression of microRNA (miR)-137 and low expression of steroid receptor coactivator-3 (SRC3) had a significant negative correlation in 40 NSCLC tissue samples. In addition, cell colony formation and proliferation was significantly reduced in miR-137-transfected A549 and NCI-H838 cells compared with scramble-transfected NSCLC cell lines. miR-137 was identified to induce G₁/S cell cycle arrest and dysregulate the mRNA expression of cell cycle-associated proteins (proliferating cell nuclear antigen, cyclin E, cyclin A1, cyclin A2 and p21) in NSCLC cells. Notably, miR-137 could significantly suppress SRC3 3' untranslated region (UTR) luciferase-reporter activity, an effect that was not detectable when the putative 3'-UTR target-site was mutated, further clarifying the molecular mechanisms underlying the role of miR-137 in NSCLC. In conclusion, the results of the present study suggest that miR-137 suppresses NSCLC cell proliferation by partially targeting SRC3.

Dysfunctional miRNA-Mediated Regulation in Chromophobe Renal Cell Carcinoma

Past research on pathogenesis of a complex disease suggests that differentially expressed message RNAs (mRNAs) can be noted as biomarkers of a disease. However, significant miRNA-mediated regulation change might also be more deep underlying cause of a disease. In this study, a miRNA-mediated regulation module is defined based on GO terms (Gene Ontology terms) from which dysfunctional modules are identified as the suspected cause of a disease. A miRNA-mediated regulation module contains mRNAs annotated to a GO term and MicroRNAs (miRNAs) which regulate the mRNAs. Based on the miRNA-mediated regulation coefficients estimated from the expression profiles of the mRNA and the miRNAs, a SW (single regulation-weight) value is then designed to evaluate the miRNA-mediated regulation change of an mRNA, and the modules with significantly differential SW values are thus identified as dysfunctional modules. The approach is applied to Chromophobe renal cell carcinoma and it identifies 70 dysfunctional miRNA-mediated regulation modules from initial 4381 modules. The identified dysfunctional modules are detected to be comprehensive reflection of chromophobe renal cell carcinoma. The proposed approach suggests that accumulated alteration in miRNA-mediated regulation might cause functional alterations, which further cause a disease. Moreover, this approach can also be used to identify diffentially miRNA-mediated regulated mRNAs showing more comprehensive underlying association with a disease than differentially expressed mRNAs.

Large-scale protein-protein interactions detection by integrating big biosensing data with computational model

Protein-protein interactions are the basis of biological functions, and studying these interactions on a molecular level is of crucial importance for understanding the functionality of a living cell. During the past decade, biosensors have emerged as an important tool for the high-throughput identification of proteins and their interactions. However, the high-throughput experimental methods for identifying PPIs are both time-consuming and expensive. On the other hand, high-throughput PPI data are often associated with high false-positive and high false-negative rates. Targeting at these problems, we propose a method for PPI detection by integrating biosensor-based PPI data with a novel computational model. This method was developed based on the algorithm of extreme learning machine combined with a novel representation of protein sequence descriptor. When performed on the large-scale human protein interaction dataset, the proposed method achieved 84.8% prediction accuracy with 84.08% sensitivity at the specificity of 85.53%. We conducted more extensive experiments to compare the proposed method with the state-of-the-art techniques, support vector machine. The achieved results demonstrate that our approach is very promising for detecting new PPIs, and it can be a helpful supplement for biosensor-based PPI data detection.

Identification of logic relationships between genes and subtypes of non-small cell lung cancer

Non-small cell lung cancer (NSCLC) has two major subtypes: adenocarcinoma (AC) and squamous cell carcinoma (SCC). The diagnosis and treatment of NSCLC are hindered by the limited knowledge about the pathogenesis mechanisms of subtypes of NSCLC. It is necessary to research the molecular mechanisms related with AC and SCC. In this work, we improved the logic analysis algorithm to mine the sufficient and necessary conditions for the presence states (presence or absence) of phenotypes. We applied our method to AC and SCC specimens, and identified lower and higher logic relationships between genes and two subtypes of NSCLC. The discovered relationships were independent of specimens selected, and their significance was validated by statistic test. Compared with the two earlier methods (the non-negative matrix factorization method and the relevance analysis method), the current method outperformed these methods in the recall rate and classification accuracy on NSCLC and normal specimens. We obtained biomarkers. Among biomarkers, genes have been used to distinguish AC from SCC in practice, and other six genes were newly discovered biomarkers for distinguishing subtypes. Furthermore, NKX2-1 has been considered as a molecular target for the targeted therapy of AC, and other genes may be novel molecular targets. By gene ontology analysis, we found that two biological processes (‘epidermis development’ and ‘cell adhesion’) were closely related with the tumorigenesis of subtypes of NSCLC. More generally, the current method could be extended to other complex diseases for distinguishing subtypes and detecting the molecular targets for targeted therapy.

Application of ThinPrep bronchial brushing cytology in the early diagnosis of lung cancer: a retrospective study

The majority of lung cancer patients are diagnosed at advanced stages of disease. This study evaluated the diagnostic value of ThinPrep (TP) bronchial brushing cytology in lung cancer. A total of 595 patients with suspicious lung cancer were enrolled in this study. The bronchial brushing samples were prepared by TP. The data were then compared to histology of lung tissue samples. Histologically, 479 of these 595 patients were diagnosed with lung cancer, including 223 cases of lung squamous cell carcinoma (SCC), 77 cases of lung adenocarcinoma (ADC), and 152 cases of small cell lung carcinoma (SCLC). The TP cytology revealed a total of 460 cases of lung cancer (including 232 SCCs, 91 ADCs, and 108 SCLCs). The TP cytological technique had 87.06% sensitivity and 62.93% specificity in the diagnosis of lung cancer. Specifically, TP cytology confirmed 195 of 223 SCCs, 47 of 77 ADCs, and 94 of 152 SCLCs. The TP cytology showed 87.44% sensitivity and 90.05% specificity for the diagnosis of SCC, with a Matthew's correlation coefficient (MCC) of 0.820; while the sensitivity was reduced to 61.04% and the specificity was 90.93% for the diagnosis of ADC, with a MCC of 0.464. For the diagnosis of SCLC, the sensitivity was 61.84% and the specificity was 96.84%, with a MCC of 0.648. Thus, this study demonstrated the usefulness of TP bronchial brushing cytology in the early diagnosis of lung cancer, especially the early stage of lung SCC. A prospective clinical trial will verify these data before being translated into the clinic.

Identification and characterization of cancer mutations in Japanese lung adenocarcinoma without sequencing of normal tissue counterparts

We analyzed whole-exome sequencing data from 97 Japanese lung adenocarcinoma patients and identified several putative cancer-related genes and pathways. Particularly, we observed that cancer-related mutation patterns were significantly different between different ethnic groups. As previously reported, mutations in the EGFR gene were characteristic to Japanese, while those in the KRAS gene were more frequent in Caucasians. Furthermore, during the course of this analysis, we found that cancer-specific somatic mutations can be detected without sequencing normal tissue counterparts. 64% of the germline variants could be excluded using a total of 217 external Japanese exome datasets. We also show that a similar approach may be used for other three ethnic groups, although the discriminative power depends on the ethnic group. We demonstrate that the ATM gene and the PAPPA2 gene could be identified as cancer prognosis related genes. By bypassing the sequencing of normal tissue counterparts, this approach provides a useful means of not only reducing the time and cost of sequencing but also analyzing archive samples, for which normal tissue counterparts are not available.