Citations

Abstract Summary: One central theme of modern high-throughput genomic data analysis is to identify relevant genomic features as well as build up a predictive model based on selected features for various tasks such as personalized medicine. Correlating the large number of ‘omics’ features with a certain phenotype is particularly challenging due to small sample size (n) and high dimensionality (p). To address this small n, large p problem, various forms of sparse regression models have been proposed by exploiting the sparsity assumption. Among these, network-constrained sparse regression model is of particular interest due to its ability to utilize the prior graph/network structure in the omics data. Despite its potential usefulness for omics data analysis, no efficient R implementation is publicly available. Here we present an R software package ‘glmgraph’ that implements the graph-constrained regularization for both sparse linear regression and sparse logistic regression. We implement both the L1 penalty and minimax concave penalty for variable selection and Laplacian penalty for coefficient smoothing. Efficient coordinate descent algorithm is used to solve the optimization problem. We demonstrate the use of the package by applying it to a human microbiome dataset, where phylogeny structure among bacterial taxa is available. Availability and implementation: ‘glmgraph’ is implemented in R and C++ Armadillo and publicly available under CRAN. Contact: chen.jun2@mayo.edu or hongzhe@upenn.edu Supplementary information: Supplementary data are available at Bioinformatics online.

Abstract Genome wide association studies (GWAS) have mapped multiple independent cancer risk loci (n = 6) to a small region on chr5p15.33 for at least ten distinct cancers, including bladder, breast, glioma, lung, melanoma, non-melanoma skin, ovarian, pancreas, prostate, and testicular germ cell cancer. This region harbors two plausible target genes, TERT which encodes the catalytic subunit of telomerase reverse transcriptase which maintains chromosome ends by adding telomeres repeats, and CLPTM1L which encodes the cleft lip and palate transmembrane protein 1-like protein which promotes cancer cell growth, protect cells from apoptosis, and can induce abnormal cytokinesis. The most significant pancreatic cancer GWAS SNP on chr5p15.33 was rs401681 (P = 3.7×10−7, OR = 1.19), located in the 13th intron of the CLPTM1L gene and within the independent multi-cancer risk locus in this region most distant from TERT. Imputation and fine mapping in pancreatic cancer resolved this signal by three orders of magnitude (P = 1.4×10−10, OR = 1.30) to a set of seven highly correlated SNPs. Electrophoretic mobility shift assays (EMSA) and luciferase assays across these SNPs led us to a single SNP, rs36115365, that showed allele-specific effects in both assays and overlapped with prominent ENCODE marks indicating regulatory potential. Examining this SNP across cell lines derived from multiple types of cancer linked to this locus, specifically pancreatic, testicular, lung, and melanoma, established a consistent pattern of a specific protein binding and enhanced regulatory activity for the risk allele across eight cancer cell lines (two cell lines per cancer type). siRNA-based targeting of the putative gene regulatory-region harboring rs36115365 resulted in a strong inhibition of TERT expression (average 60%, range 45-85%) but no effect on CLPTM1L expression. This inhibition of TERT expression was stronger from the pancreatic cancer risk as compared to the protective allele. Quantitative mass spectrometry identified Zinc finger protein 148 (ZNF148) as the protein specifically binding the risk allele, further supported by direct and allele-specific binding of purified recombinant ZNF148 in EMSA assays. Lastly, siRNA-mediated knockdown of ZNF148 resulted in significant reduction of TERT but not CLPTM1L expression across multiple cancer cell lines. Our results indicate that the pancreatic cancer GWAS signal at the single independent chr5p15.33 multi-cancer risk locus most distant from TERT may be explained by a single SNP that influences TERT expression via ZNF148. Citation Format: Kevin M. Brown, Jun Fang, Jinping Jia, Zhaoming Wang, Matthew Makowski, Tongwu Zhang, Jason Hoskins, Jiyeon Choi, Younghun Han, Mingfeng Zhang, Mai Xu, Peter Kanetsky, Andresson Thorkell, Gloria M. Petersen, Katherine L. Nathanson, Christopher I. Amos, Maria T. Landi, Stephen J. Chanock, Michiel Vermeulen, Laufey T. Amundadottir. Functional characterization of a multicancer risk locus on chr5p15.33 reveals regulation of TERT by ZNF148. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4610. doi:10.1158/1538-7445.AM2015-4610

AbstractWe describe here the vision, motivations, and research plans of the National Institutes of Health Center for Excellence in Big Data Computing at the University of Illinois, Urbana-Champaign. The Center is organized around the construction of “Knowledge Engine for Genomics” (KnowEnG), an E-science framework for genomics where biomedical scientists will have access to powerful methods of data mining, network mining, and machine learning to extract knowledge out of genomics data. The scientist will come to KnowEnG with their own data sets in the form of spreadsheets and ask KnowEnG to analyze those data sets in the light of a massive knowledge base of community data sets called the “Knowledge Network” that will be at the heart of the system. The Center is undertaking discovery projects aimed at testing the utility of KnowEnG for transforming big data to knowledge. These projects span a broad range of biological enquiry, from pharmacogenomics (in collaboration with Mayo Clinic) to transcriptomics of human behavior.

Trafficking from bedside to bench Typically in translational research, a discovery in cell or molecular biology is later exploited to improve patient care. Occasionally, information flows in the opposite direction. Lo et al. found that patients with an autoimmune disorder caused by deficiency of a protein called LRBA responded dramatically to the drug abatacept (see the Perspective by Sansom). Abatacept contains a segment of a potent inhibitory immune receptor, CTLA4. Experiments prompted by this observation revealed the relationship between the two proteins: LRBA controls the intracellular trafficking and degradation of CTLA4. This information may further improve patient care, because other clinically approved drugs have the desired mechanism of action with potentially fewer side effects. Science , this issue p. 436 ; see also p. 377

Background— Apixaban is approved for the prevention of stroke and systemic embolism in patients with nonvalvular atrial fibrillation. However, the Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation (ARISTOTLE) trial included a substantial number of patients with valvular heart disease and only excluded patients with clinically significant mitral stenosis or mechanical prosthetic heart valves. Methods and Results— We compared the effect of apixaban and warfarin on rates of stroke or systemic embolism, major bleeding, and death in patients with and without moderate or severe valvular heart disease using Cox proportional hazards modeling. Of the 18 201 patients enrolled in ARISTOTLE, 4808 (26.4%) had a history of moderate or severe valvular heart disease or previous valve surgery. Patients with valvular heart disease had higher rates of stroke or systemic embolism and bleeding than patients without valvular heart disease. There was no evidence of a differential effect of apixaban over warfarin in patients with and without valvular heart disease in reducing stroke and systemic embolism (hazard ratio [HR], 0.70; 95% confidence interval [CI], 0.51–0.97 and HR, 0.84; 95%, CI 0.67–1.04; interaction P =0.38), causing less major bleeding (HR, 0.79; 95% CI, 0.61–1.04 and HR, 0.65; 95% CI, 0.55–0.77; interaction P =0.23), and reducing mortality (HR, 1.01; 95% CI, 0.84–1.22 and HR, 0.84; 95% CI, 0.73–0.96; interaction P =0.10). Conclusions— More than a quarter of the patients in ARISTOTLE with nonvalvular atrial fibrillation had moderate or severe valvular heart disease. There was no evidence of a differential effect of apixaban over warfarin in reducing stroke or systemic embolism, causing less bleeding, and reducing death in patients with and without valvular heart disease. Clinical Trial Registration— URL: http://www.clinicaltrials.gov . Unique identifier: NCT00412984.

AbstractIntroductionAfrican‐American (AA) individuals have a higher risk for late‐onset Alzheimer's disease (LOAD) than Americans of primarily European ancestry (EA). Recently, the largest genome‐wide association study in AAs to date confirmed that six of the Alzheimer's disease (AD)‐related genetic variants originally discovered in EA cohorts are also risk variants in AA; however, the risk attributable to many of the loci (e.g., APOE, ABCA7) differed substantially from previous studies in EA. There likely are risk variants of higher frequency in AAs that have not been discovered.MethodsWe performed a comprehensive analysis of genetically determined local and global ancestry in AAs with regard to LOAD status.ResultsCompared to controls, LOAD cases showed higher levels of African ancestry, both globally and at several LOAD relevant loci, which explained risk for AD beyond global differences.DiscussionExploratory post hoc analyses highlight regions with greatest differences in ancestry as potential candidate regions for future genetic analyses.

Abstract A genome-wide association study (GWAS) of prostate cancer in Kaiser Permanente health plan members (7,783 cases, 38,595 controls; 80.3% non-Hispanic white, 4.9% African-American, 7.0% East Asian, and 7.8% Latino) revealed a new independent risk indel rs4646284 at the previously identified locus 6q25.3 that replicated in PEGASUS (N = 7,539) and the Multiethnic Cohort (N = 4,679) with an overall P = 1.0 × 10−19 (OR, 1.18). Across the 6q25.3 locus, rs4646284 exhibited the strongest association with expression of SLC22A1 (P = 1.3 × 10−23) and SLC22A3 (P = 3.2 × 10−52). At the known 19q13.33 locus, rs2659124 (P = 1.3 × 10−13; OR, 1.18) nominally replicated in PEGASUS. A risk score of 105 known risk SNPs was strongly associated with prostate cancer (P < 1.0 × 10−8). Comparing the highest to lowest risk score deciles, the OR was 6.22 for non-Hispanic whites, 5.82 for Latinos, 3.77 for African-Americans, and 3.38 for East Asians. In non-Hispanic whites, the 105 risk SNPs explained approximately 7.6% of disease heritability. The entire GWAS array explained approximately 33.4% of heritability, with a 4.3-fold enrichment within DNaseI hypersensitivity sites (P = 0.004). Significance: Taken together, our findings of independent risk variants, ethnic variation in existing SNP replication, and remaining unexplained heritability have important implications for further clarifying the genetic risk of prostate cancer. Our findings also suggest that there may be much promise in evaluating understudied variation, such as indels and ethnically diverse populations. Cancer Discov; 5(8); 878–91. ©2015 AACR. This article is highlighted in the In This Issue feature, p. 783

BNLF2a is an Epstein‐Barr virus (EBV) immune evasion gene. Its protein is located in the endoplasmic reticulum (ER) membrane, and can inhibit the antigen transporting function of TAP, thereby perturbing the immune response to EBV in lytic and prelatent phase. In order to explore whether the polymorphism of BNLF2a gene has a role in different types of EBV associated tumors, we conducted complete sequencing of the gene BNLF2a in 408 cases of EBV positive tumors (76 lymphomas, 45 gastric carcinomas, and 85 nasopharyngeal carcinomas in northern China and 27 lymphomas, 30 gastric carcinomas, and 57 nasopharyngeal carcinomas in southern China) and throat washings from healthy individuals (39 in northern China and 49 in southern China). Two main variant types of BNLF2a were identified. Type BNLF2a‐A, which was similar to B95‐8, was dominant in all sub‐populations (66.7–100%) in this study. Type BNLF2a‐B was characterized by the mutations at position 8 and 40. The variation patterns of BNLF2a were significantly different between samples from northern and southern China ( P < 0.05), and between the tumors and healthy donor samples from the northern China ( P < 0.0167). Type BNLF2a‐B was more frequent in healthy donors of northern China (33.3%), and the proportion of this type was higher in the northern than in the southern NPCs. These data demonstrate that the BNLF2a gene is highly conserved, and its polymorphism is geographically restricted. Type BNLF2a‐B is more prevalent in northern China and may be less tumor transformative. J. Med. Virol. 87:1946–1952, 2015 . © 2015 Wiley Periodicals, Inc.