Optical Genome Mapping Reveals the Landscape of Structural Variations and Their Clinical Significance in HBOC-Related Breast Cancer.

BACKGROUND: Structural variations (SVs) are common genetic alterations in the human genome. However, the profile and clinical relevance of SVs in patients with hereditary breast and ovarian cancer (HBOC) syndrome (germline BRCA1/2 mutations) remains to be fully elucidated. METHODS: Twenty HBOC-related cancer samples (5 breast and 15 ovarian cancers) were studied by optical genome mapping (OGM) and next-generation sequencing (NGS) assays. RESULTS: The SV landscape in the 5 HBOC-related breast cancer samples was comprehensively investigated to determine the impact of intratumor SV heterogeneity on clinicopathological features and on the pattern of genetic alteration. SVs and copy number variations (CNVs) were common genetic events in HBOC-related breast cancer, with a median of 212 SVs and 107 CNVs per sample. The most frequently detected type of SV was insertion, followed by deletion. The 5 HBOC-related breast cancer samples were divided into SVhigh and SVlow groups according to the intratumor heterogeneity of SVs. SVhigh tumors were associated with higher Ki-67 expression, higher homologous recombination deficiency (HRD) scores, more mutated genes, and altered signaling pathways. Moreover, 60% of the HBOC-related breast cancer samples displayed chromothripsis, and 8 novel gene fusion events were identified by OGM and validated by transcriptome data. CONCLUSIONS: These findings suggest that OGM is a promising tool for the detection of SVs and CNVs in HBOC-related breast cancer. Furthermore, OGM can efficiently characterize chromothripsis events and novel gene fusions. SVhigh HBOC-related breast cancers were associated with unfavorable clinicopathological features. SVs may therefore have predictive and therapeutic significance for HBOC-related breast cancers in the clinic.

Dominant neoantigen verification in hepatocellular carcinoma by a single-plasmid system coexpressing patient HLA and antigen.

BACKGROUND: Previous studies confirmed that most neoantigens predicted by algorithms do not work in clinical practice, and experimental validations remain indispensable for confirming immunogenic neoantigens. In this study, we identified the potential neoantigens with tetramer staining, and established the Co-HA system, a single-plasmid system coexpressing patient human leukocyte antigen (HLA) and antigen, to detect the immunogenicity of neoantigens and verify new dominant hepatocellular carcinoma (HCC) neoantigens. METHODS: First, we enrolled 14 patients with HCC for next-generation sequencing for variation calling and predicting potential neoantigens. Then, the Co-HA system was established. To test the feasibility of the system, we constructed target cells coexpressing HLA-A*11:01 and the reported KRAS G12D neoantigen as well as specific T-cell receptor (TCR)-T cells. The specific cytotoxicity generated by this neoantigen was shown using the Co-HA system. Moreover, potential HCC-dominant neoantigens were screened out by tetramer staining and validated by the Co-HA system using methods including flow cytometry, enzyme-linked immunospot assay and ELISA. Finally, antitumor test in mouse mode and TCR sequencing were performed to further evaluate the dominant neoantigen. RESULTS: First, 2875 somatic mutations in 14 patients with HCC were identified. The main base substitutions were C>T/G>A transitions, and the main mutational signatures were 4, 1 and 16. The high-frequency mutated genes included HMCN1, TTN and TP53. Then, 541 potential neoantigens were predicted. Importantly, 19 of the 23 potential neoantigens in tumor tissues also existed in portal vein tumor thrombi. Moreover, 37 predicted neoantigens restricted by HLA-A*11:01, HLA-A*24:02 or HLA-A*02:01 were performed by tetramer staining to screen out potential HCC-dominant neoantigens. HLA-A*24:02-restricted epitope 5'-FYAFSCYYDL-3' and HLA-A*02:01-restricted epitope 5'-WVWCMSPTI-3' demonstrated strong immunogenicity in HCC, as verified by the Co-HA system. Finally, the antitumor efficacy of 5'-FYAFSCYYDL-3'-specific T cells was verified in the B-NDG-B2mtm1Fcrntm1(mB2m) mouse and their specific TCRs were successfully identified. CONCLUSION: We found the dominant neoantigens with high immunogenicity in HCC, which were verified with the Co-HA system.

Drug screening and biomarker gene investigation in cancer therapy through the human transcriptional regulatory network.

A complex and vast biological network regulates all biological functions in the human body in a sophisticated manner, and abnormalities in this network can lead to disease and even cancer. The construction of a high-quality human molecular interaction network is possible with the development of experimental techniques that facilitate the interpretation of the mechanisms of drug treatment for cancer. We collected 11 molecular interaction databases based on experimental sources and constructed a human protein-protein interaction (PPI) network and a human transcriptional regulatory network (HTRN). A random walk-based graph embedding method was used to calculate the diffusion profiles of drugs and cancers, and a pipeline was constructed by using five similarity comparison metrics combined with a rank aggregation algorithm, which can be implemented for drug screening and biomarker gene prediction. Taking NSCLC as an example, curcumin was identified as a potentially promising anticancer drug from 5450 natural small molecules, and combined with differentially expressed genes, survival analysis, and topological ranking, we obtained BIRC5 (survivin), which is both a biomarker for NSCLC and a key target for curcumin. Finally, the binding mode of curcumin and survivin was explored using molecular docking. This work has a guiding significance for antitumor drug screening and the identification of tumor markers.

Biological knowledge graph-guided investigation of immune therapy response in cancer with graph neural network.

The determination of transcriptome profiles that mediate immune therapy in cancer remains a major clinical and biological challenge. Despite responses induced by immune-check points inhibitors (ICIs) in diverse tumor types and all the big breakthroughs in cancer immunotherapy, most patients with solid tumors do not respond to ICI therapies. It still remains a big challenge to predict the ICI treatment response. Here, we propose a framework with multiple prior knowledge networks guided for immune checkpoints inhibitors prediction-DeepOmix-ICI (or ICInet for short). ICInet can predict the immune therapy response by leveraging geometric deep learning and prior biological knowledge graphs of gene-gene interactions. Here, we demonstrate more than 600 ICI-treated patients with ICI response data and gene expression profile to apply on ICInet. ICInet was used for ICI therapy responses prediciton across different cancer types-melanoma, gastric cancer and bladder cancer, which includes 7 cohorts from different data sources. ICInet is able to robustly generalize into multiple cancer types. Moreover, the performance of ICInet in those cancer types can outperform other ICI biomarkers in the clinic. Our model [area under the curve (AUC = 0.85)] generally outperformed other measures, including tumor mutational burden (AUC = 0.62) and programmed cell death ligand-1 score (AUC = 0.74). Therefore, our study presents a prior-knowledge guided deep learning method to effectively select immunotherapy-response-associated biomarkers, thereby improving the prediction of immunotherapy response for precision oncology.

Genomic and transcriptomic profiling indicates the prognosis significance of mutational signature for TMB-high subtype in Chinese patients with gastric cancer.

INTRODUCTION: Gastric cancer (GC)is the third leading cause of cancer-related deaths in China and immunotherapy emerging as a revolutionary treatment for GC recently. Tumor mutational burden (TMB) is a predictive biomarker of immunotherapy in multiple cancers. However, the prognostic significance and subtype of TMB in GC is not fully understood. OBJECTIVES: This study aims to evaluate the prognostic value of TMB in Chinese GC and further classify TMB-high GC (GCTMB-H) patients combing with mutational signatures. METHODS: Genomic profiling of 435 cancer-gene panel was performed using 206 GC samples from Chinese people. Actionable genetic alterations were compared across all the samples to generate actionable subtyping. The prognostic value of TMB in Chinese GC was evaluated. Mutational signatures were analyzed on TMB-H subtype to stratify the prognosis of TMB. Transcriptomic analysis was applied to compare the distributed immunocytes among different subtypes. RESULTS: 88.3% (182/206) of GC samples had at least one mutation, while 45.1% (93/206) had at least one somatic copy number alteration (SCNA). 29.6% (61/206) of GC samples were TMB-H, including 13 MSI-H and 48 MSS tumors. According to distinct genetic alteration profiles of 69 actionable genes, we classified GC samples into eight molecular subtypes, including TMB-H, ERBB2 amplified, ATM mutated, BRCA2 mutated, CDKN2A/B deleted, PI3KCA mutated, KRAS mutated, and less-mutated subtype. TMB-H subtype presented a remarkable immune-activated phenotype as determined by transcriptomic analysis that was further validated in the TCGA GC cohort. GCTMB-H patients exhibited significantly better survival (P = 0.047). But Signature 1-high GCTMB-H patients had relatively worse prognosis (P = 0.0209, HR = 2.571) than Signature 1-low GCTMB-H patients from Chinese GC cohort, also validated in TCGA GC cohort, presenting highly activated carbohydrate, fatty acid or lipid metabolism. CONCLUSION: The Signature 1-high GCTMB-H could be a marker of poor prognosis and is associated with metabolism disorder.

High Throughput Transcriptome Data Analysis and Computational Verification Reveal Immunotherapy Biomarkers of Compound Kushen Injection for Treating Triple-Negative Breast Cancer.

BACKGROUND: Although notable therapeutic and prognostic benefits of compound kushen injection (CKI) have been found when it was used alone or in combination with chemotherapy or radiotherapy for triple-negative breast cancer (TNBC) treatment, the effects of CKI on TNBC microenvironment remain largely unclear. This study aims to construct and validate a predictive immunotherapy signature of CKI on TNBC. METHODS: The UPLC-Q-TOF-MS technology was firstly used to investigate major constituents of CKI. RNA sequencing data of CKI-perturbed TNBC cells were analyzed to detect differential expression genes (DEGs), and the GSVA algorithm was applied to explore significantly changed pathways regulated by CKI. Additionally, the ssGSEA algorithm was used to quantify immune cell abundance in TNBC patients, and these patients were classified into distinct immune infiltration subgroups by unsupervised clustering. Then, prognosis-related genes were screened from DEGs among these subgroups and were further overlapped with the DEGs regulated by CKI. Finally, a predictive immunotherapy signature of CKI on TNBC was constructed based on the LASSO regression algorithm to predict mortality risks of TNBC patients, and the signature was also validated in another TNBC cohort. RESULTS: Twenty-three chemical components in CKI were identified by UPLC-Q-TOF-MS analysis. A total of 3692 DEGs were detected in CKI-treated versus control groups, and CKI significantly activated biological processes associated with activation of T, natural killer and natural killer T cells. Three immune cell infiltration subgroups with 1593 DEGs were identified in TNBC patients. Then, two genes that can be down-regulated by CKI with hazard ratio (HR) > 1 and 26 genes that can be up-regulated by CKI with HR < 1 were selected as key immune- and prognosis-related genes regulated by CKI. Lastly, a five-gene prognostic signature comprising two risky genes (MARVELD2 and DYNC2I2) that can be down-regulated by CKI and three protective genes (RASSF2, FERMT3 and RASSF5) that can be up-regulated by CKI was developed, and it showed a good performance in both training and test sets. CONCLUSIONS: This study proposes a predictive immunotherapy signature of CKI on TNBC, which would provide more evidence for survival prediction and treatment guidance in TNBC as well as a paradigm for exploring immunotherapy biomarkers in compound medicines.

Single-cell Long Non-coding RNA Landscape of T Cells in Human Cancer Immunity.

The development of new biomarkers or therapeutic targets for cancer immunotherapies requires deep understanding of T cells. To date, the complete landscape and systematic characterization of long noncoding RNAs (lncRNAs) in T cells in cancer immunity are lacking. Here, by systematically analyzing full-length single-cell RNA sequencing (scRNA-seq) data of more than 20,000 libraries of T cells across three cancer types, we provided the first comprehensive catalog and the functional repertoires of lncRNAs in human T cells. Specifically, we developed a custom pipeline for de novotranscriptome assembly and obtained a novel lncRNA catalog containing 9433 genes. This increased the number of current human lncRNA catalog by 16% and nearly doubled the number of lncRNAs expressed in T cells. We found that a portion of expressed genes in single T cells were lncRNAs which had been overlooked by the majority of previous studies. Based on metacell maps constructed by the MetaCell algorithm that partitions scRNA-seq datasets into disjointed and homogenous groups of cells (metacells), 154 signature lncRNA genes were identified. They were associated with effector, exhausted, and regulatory T cell states. Moreover, 84 of them were functionally annotated based on the co-expression networks, indicating that lncRNAs might broadly participate in the regulation of T cell functions. Our findings provide a new point of view and resource for investigating the mechanisms of T cell regulation in cancer immunity as well as for novel cancer-immune biomarker development and cancer immunotherapies.

DeepOmix: A scalable and interpretable multi-omics deep learning framework and application in cancer survival analysis.

Integrative analysis of multi-omics data can elucidate valuable insights into complex molecular mechanisms for various diseases. However, due to their different modalities and high dimension, utilizing and integrating different types of omics data suffers from great challenges. There is an urgent need to develop a powerful method to improve survival prediction and detect functional gene modules from multi-omics data. To deal with these problems, we present DeepOmix (a scalable and interpretable multi-Omics Deep learning framework and application in cancer survival analysis), a flexible, scalable, and interpretable method for extracting relationships between the clinical survival time and multi-omics data based on a deep learning framework. DeepOmix enables the non-linear combination of variables from different omics datasets and incorporates prior biological information defined by users (such as signaling pathways and tissue networks). Benchmark experiments demonstrate that DeepOmix outperforms the other five cutting-edge prediction methods. Besides, Lower Grade Glioma (LGG) is taken as the case study to perform the prognosis prediction and illustrate the functional module nodes which are associated with the prognostic result in the prediction model.

Imbalance of TGF-ß1/BMP-7 pathways induced by M2-polarized macrophages promotes hepatocellular carcinoma aggressiveness.

The transforming growth factor-beta (TGF-ß) signaling pathway is the predominant cytokine signaling pathway in the development and progression of hepatocellular carcinoma (HCC). Bone morphogenetic protein (BMP), another member of the TGF-ß superfamily, has been frequently found to participate in crosstalk with the TGF-ß pathway. However, the complex interaction between the TGF-ß and BMP pathways has not been fully elucidated in HCC. We found that the imbalance of TGF-ß1/BMP-7 pathways was associated with aggressive pathological features and poor clinical outcomes in HCC. The induction of the imbalance of TGF-ß1/BMP-7 pathways in HCC cells could significantly promote HCC cell invasion and stemness by increasing inhibitor of differentiation 1 (ID1) expression. We also found that the microRNA (miR)-17-92 cluster, originating from the extracellular vesicles (EVs) of M2-polarized tumor-associated macrophages (M2-TAMs), stimulated the imbalance of TGF-ß1/BMP-7 pathways in HCC cells by inducing TGF-ß type II receptor (TGFBR2) post-transcriptional silencing and inhibiting activin A receptor type 1 (ACVR1) post-translational ubiquitylation by targeting Smad ubiquitylation regulatory factor 1 (Smurf1). In vivo, short hairpin (sh)-MIR17HG and ACVR1 inhibitors profoundly attenuated HCC cell growth and metastasis by rectifying the imbalance of TGF-ß1/BMP-7 pathways. Therefore, we proposed that the imbalance of TGF-ß1/BMP-7 pathways is a feasible prognostic biomarker and recovering the imbalance of TGF-ß1/BMP-7 pathways might be a potential therapeutic strategy for HCC.

VarBen: Generating in Silico Reference Data Sets for Clinical Next-Generation Sequencing Bioinformatics Pipeline Evaluation.

Next-generation sequencing is increasingly being adopted as a valuable method for the detection of somatic variants in clinical oncology. However, it is still challenging to reach a satisfactory level of robustness and standardization in clinical practice when using the currently available bioinformatics pipelines to detect variants from raw sequencing data. Moreover, appropriate reference data sets are lacking for clinical bioinformatics pipeline development, validation, and proficiency testing. Herein, we developed the Variant Benchmark tool (VarBen), an open-source software for variant simulation to generate customized reference data sets by directly editing the original sequencing reads. VarBen can introduce a variety of variants, including single-nucleotide variants, small insertions and deletions, and large structural variants, into targeted, exome, or whole-genome sequencing data, and can handle sequencing data from both the Illumina and Ion Torrent sequencing platforms. To demonstrate the feasibility and robustness of VarBen, we performed variant simulation on different sequencing data sets and compared the simulated variants with real-world data. The validation study showed that the simulated data are highly comparable to real-world data and that VarBen is a reliable tool for variant simulation. In addition, our collaborative study of somatic variant calling in 20 laboratories emphasizes the need for laboratories to evaluate their bioinformatics pipelines with customized reference data sets. VarBen may help users develop and validate their bioinformatics pipelines using locally generated sequencing data.

NONCODEV6: an updated database dedicated to long non-coding RNA annotation in both animals and plants.

NONCODE (http://www.noncode.org/) is a comprehensive database of collection and annotation of noncoding RNAs, especially long non-coding RNAs (lncRNAs) in animals. NONCODEV6 is dedicated to providing the full scope of lncRNAs across plants and animals. The number of lncRNAs in NONCODEV6 has increased from 548 640 to 644 510 since the last update in 2017. The number of human lncRNAs has increased from 172 216 to 173 112. The number of mouse lncRNAs increased from 131 697 to 131 974. The number of plant lncRNAs is 94 697. The relationship between lncRNAs in human and cancer were updated with transcriptome sequencing profiles. Three important new features were also introduced in NONCODEV6: (i) updated human lncRNA-disease relationships, especially cancer; (ii) lncRNA annotations with tissue expression profiles and predicted function in five common plants; iii) lncRNAs conservation annotation at transcript level for 23 plant species. NONCODEV6 is accessible through http://www.noncode.org/.

Phenotype-Genotype Association Analysis of ACTH-Secreting Pituitary Adenoma and Its Molecular Link to Patient Osteoporosis.

Adrenocorticotrophin (ACTH)-secreting pituitary adenoma, also known as Cushing disease (CD), is rare and causes metabolic syndrome, cardiovascular disease and osteoporosis due to hypercortisolism. However, the molecular pathogenesis of CD is still unclear because of a lack of human cell lines and animal models. Here, we study 106 clinical characteristics and gene expression changes from 118 patients, the largest cohort of CD in a single-center. RNA deep sequencing is used to examine genotypic changes in nine paired female ACTH-secreting pituitary adenomas and adjacent nontumorous pituitary tissues (ANPT). We develop a novel analysis linking disease clinical characteristics and whole transcriptomic changes, using Pearson Correlation Coefficient to discover a molecular network mechanism. We report that osteoporosis is distinguished from the phenotype and genotype analysis. A cluster of genes involved in osteoporosis is identified using Pearson correlation coefficient analysis. Most of the genes are reported in the bone related literature, confirming the feasibility of phenotype-genotype association analysis, which could be used in the analysis of almost all diseases. Secreted phosphoprotein 1 (SPP1), collagen type I α 1 chain (COL1A1), 5'-nucleotidase ecto (NT5E), HtrA serine peptidase 1 (HTRA1) and angiopoietin 1 (ANGPT1) and their signalling pathways are shown to be involved in osteoporosis in CD patients. Our discoveries provide a molecular link for osteoporosis in CD patients, and may open new potential avenues for osteoporosis intervention and treatment.

Identification of prognostic biomarkers in hepatitis B virus-related hepatocellular carcinoma and stratification by integrative multi-omics analysis.

BACKGROUND & AIMS: The differentiation of distinct multifocal hepatocellular carcinoma (HCC): multicentric disease vs. intrahepatic metastases, in which the management and prognosis varies substantively, remains problematic. We aim to stratify multifocal HCC and identify novel diagnostic and prognostic biomarkers by performing whole genome and transcriptome sequencing, as part of a multi-omics strategy. METHODS: A complete collection of tumour and somatic specimens (intrahepatic HCC lesions, matched non-cancerous liver tissue and blood) were obtained from representative patients with multifocal HCC exhibiting two distinct postsurgical courses. Whole-genome and transcriptome sequencing with genotyping were performed for each tissue specimen to contrast genomic alterations, including hepatitis B virus integrations, somatic mutations, copy number variations, and structural variations. We then constructed a phylogenetic tree to visualise individual tumour evolution and performed functional enrichment analyses on select differentially expressed genes to elucidate biological processes involved in multifocal HCC development. Multi-omics data were integrated with detailed clinicopathological information to identify HCC biomarkers, which were further validated using a large cohort of HCC patients (n = 174). RESULTS: The multi-omics profiling and tumour biomarkers could successfully distinguish the two multifocal HCC types, while accurately predicting clonality and aggressiveness. The dual-specificity protein kinase TTK, which is a key mitotic checkpoint regulator with links to p53 signaling, was further shown to be a promising overall prognostic marker for HCC in the large patient cohort. CONCLUSIONS: Comprehensive multi-omics characterisation of multifocal tumour evolution may improve clinical decision-making, facilitate personalised medicine, and expedite identification of novel biomarkers and therapeutic targets in HCC.

Downregulation of Insulin-like growth factor binding protein 6 is associated with ACTH-secreting pituitary adenoma growth.

BACKGROUND: Adrenocorticotrophic hormone (ACTH)-dependent Cushing's syndrome, called Cushing disease, is caused by a corticotroph tumor of the pituitary gland. Insulin-like growth factor binding protein 6 (IGFBP6), which regulates insulin-like growth factor (IGF) activity and inhibits several IGF2-dependent cancer growths, plays a pivotal role in the tumorigenesis of malignancy, but its roles in ACTH-secreting pituitary adenomas remain unclear. OBJECTIVE: To investigate IGFBP6 expression in ACTH-secreting pituitary adenomas, and its involvement in tumor growth. METHODS: Sporadic ACTH-secreting pituitary adenomas specimens (n = 41) and adjacent non-tumorous pituitary tissues (n = 9) were collected by transphenoidal surgery. IGFBP6 expression was assessed by quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) and validated by Western blotting. Associations of IGFBP6 expression with maximum tumor diameter or Ki-67 labeling index were evaluated in ACTH-secreting pituitary adenomas. RESULTS: IGFBP6 mRNA and protein expression were both decreased in ACTH-secreting pituitary adenomas, compared to adjacent non-tumorous pituitary tissues (P < 0.01). IGFBP6 expression was correlated inversely with maximum tumor diameter (Rho = -0.53, P < 0.0001) and Ki-67 levels (Rho = -0.52, P < 0.05). Moreover, IGFBP6 downregulation activated PI3 K-AKT-mTOR pathway in ACTH-secreting pituitary adenomas. CONCLUSIONS: IGFBP6 attenuation in ACTH-secreting pituitary adenomas is associated with tumor growth, through activation of PI3K-AKT-mTOR pathway. The finding underlies IGFBP6 roles in Cushing disease and would potentially provide a novel target of medical therapies.

Genome-wide identification of cancer-related polyadenylated and non-polyadenylated RNAs in human breast and lung cell lines.

Eukaryotic mRNAs consist of two forms of transcripts: poly(A)+ and poly(A)-, based on the presence or absence of poly(A) tails at the 3' end. Poly(A)+ mRNAs are mainly protein coding mRNAs, whereas the functions of poly(A)- mRNA are largely unknown. Previous studies have shown that a significant proportion of gene transcripts are poly(A)- or bimorphic (containing both poly(A)+ and poly(A)- transcripts). We compared the expression levels of poly(A)- and poly(A)+ RNA mRNAs in normal and cancer cell lines. We also investigated the potential functions of these RNA transcripts using an integrative workflow to explore poly(A)+ and poly(A)- transcriptome sequences between a normal human mammary gland cell line (HMEC) and a breast cancer cell line (MCF-7), as well as between a normal human lung cell line (NHLF) and a lung cancer cell line (A549). The data showed that normal and cancer cell lines differentially express these two forms of mRNA. Gene ontology (GO) annotation analyses hinted at the functions of these two groups of transcripts and grouped the differentially expressed genes according to the form of their transcript. The data showed that cell cycle-, apoptosis-, and cell death-related functions corresponded to most of the differentially expressed genes in these two forms of transcripts, which were also associated with the cancers. Furthermore, translational elongation and translation functions were also found for the poly(A)- protein-coding genes in cancer cell lines. We demonstrate that poly(A)- transcripts play an important role in cancer development.

Systematic study of human long intergenic non-coding RNAs and their impact on cancer.

The functional impact of several long intergenic non-coding RNAs (lincRNAs) has been characterized in previous studies. However, it is difficult to identify lincRNAs on a large-scale and to ascertain their functions or predict their structures in laboratory experiments because of the diversity, lack of knowledge and specificity of expression of lincRNAs. Furthermore, although there are a few well-characterized examples of lincRNAs associated with cancers, these are just the tip of the iceberg owing to the complexity of cancer. Here, by combining RNA-Seq data from several kinds of human cell lines with chromatin-state maps and human expressed sequence tags, we successfully identified more than 3000 human lincRNAs, most of which were new ones. Subsequently, we predicted the functions of 105 lincRNAs based on a coding-non-coding gene co-expression network. Finally, we propose a genetic mediator and key regulator model to unveil the subtle relationships between lincRNAs and lung cancer. Twelve lincRNAs may be principal players in lung tumorigenesis. The present study combines large-scale identification and functional prediction of human lincRNAs, and is a pioneering work in characterizing cancer-associated lincRNAs by bioinformatics.