DIVAN: accurate identification of non-coding disease-specific risk variants using multi-omics profiles

Abstract

Understanding the link between non-coding sequence variants, identified in genome-wide association studies, and the pathophysiology of complex diseases remains challenging due to a lack of annotations in non-coding regions. To overcome this, we developed DIVAN, a novel feature selection and ensemble learning framework, which identifies disease-specific risk variants by leveraging a comprehensive collection of genome-wide epigenomic profiles across cell types and factors, along with other static genomic features. DIVAN accurately and robustly recognizes non-coding disease-specific risk variants under multiple testing scenarios; among all the features, histone marks, especially those marks associated with repressed chromatin, are often more informative than others.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-016-1112-z) contains supplementary material, which is available to authorized users.

IW-Scoring: an Integrative Weighted Scoring framework for annotating and prioritizing genetic variations in the noncoding genome

“…However, IW-Scoring still allows for the functional variants associated with specific tissues, cells and features to be identified through the regulatory annotation module. This is currently lacking in many other methods, although some algorithms have chosen to focus on the identification of disease/tissue specific risk variants recently [22, 41]. Compared to most available methods, we believe our approach is optimally balanced between summarised and detailed evidences for the diverse range of users.…”

“…Via a vigorous weight learning process, strong weights were assigned to the block of closely correlated scores (Eigen, DeepSEA, FATHMM noncoding, ReMM and CADD), and the derived IW-Scoring significantly outperformed individual constituent scores (including Eigen and Eigen-PC) across various data sets, demonstrating the accuracy and validity of our approach. Such ensemble based approach with different estimated weights has been shown to perform better than any single component classifier [26], and has been widely used in various bioinformatics problems [41, 42]. The weighted integration technique based on the eigendecomposition of the covariance matrix also offers the flexibility to incorporate any other correlated genome-wide functional scores/features into the integrative scores.…”

Detailed Regulatory Interaction Map of the Human Heart Facilitates Gene Discovery for Cardiovascular Disease

“…Such variants do not change the functionality of a gene product but instead are thought to affect transcript levels through distant gene regulation. In support of this concept, many of these genomic sites display transcription regulatory potential in disease-relevant cell-types, as judged from their epigenetic signatures (1000Genomes Project Consortium et al, 2010Chen et al, 2016;Maurano et al, 2012). Gene regulation over distance requires chromatin looping to bring distal regulatory DNA elements in close proximity to target genes.…”

“…However, the identification of target genes and underlying mechanisms remains challenging. Indeed, more than 90% of disease-associated variation resides in noncoding DNA (1000Genomes Project Consortium et al, 2010Chen et al, 2016;Maurano et al, 2012).…”

IPEV: a web server for inferring pathogenic enhancers with variants

“…Feature selection and outlier removal were employed to achieve the best performance. The optimal feature set was selected depend on the largest area under the receiver operating characteristic curve (ROC-AUC) value as described in previous study [27]. Briefly, the confidence of each feature was measured by p values based on Wilcoxon rank sum test.…”