pVAAST: A new method for family-based rare variant association testing. C. D. Huff1, H. Hu1, H. Coon2, S. Guthery2, S. Tavtigian2, J. C. Roach3, Z. Kronenberg2, J. Xing4, G. Glusman3, V. Garg6, B. Moore2, L. E. Hood3, K. S. Pollard5, D. J. Galas7, D. Srivastava5, M. G. Reese8, L. B. Jorde2, M. Yandell2 1) Epidemiology, The University of Texas M.D. Anderson Cancer Center, Houston, TX; 2) University of Utah School of Medicine, Salt Lake City, Utah, USA; 3) Institute for Systems Biology, Seattle, WA, USA; 4) Department of Genetics, Rutgers University, Piscataway, NJ, USA; 5) Gladstone Institute of Cardiovascular Disease and University of California, San Francisco, San Francisco, CA, USA; 6) Department of Pediatrics, The Ohio State University and Center for Cardiovascular and Pulmonary Research, Research Institute at Nationwide Children's Hospital, Columbus, OH, USA; 7) Luxembourg Centre for Systems Biomedicine, University of Luxembourg, Esch-sur-Alzette, Luxembourg; 8) Omicia, Inc., Emeryville, CA, USA.
Next-generation sequencing has become an important tool for identifying disease-causing variation in families. Although a variety of filtering methods have been successfully applied in family-based sequencing studies, few robust statistical methods are available to support such studies. The Variant Annotation, Analysis and Search Tool (VAAST) employs a variant association test that combines amino acid substitution and allele frequency information using a composite likelihood ratio test (CLRT). Here, we present a novel pedigree-based method, pedigree-VAAST (pVAAST), that expands VAAST to incorporate family data. The method evaluates the familial evidence using a model specifically designed for sequence data. This model is broadly similar to traditional linkage analysis but is more sensitive when the disease alleles are modestly rare (minor allele frequency 0.01-0.05), which is a critical parameter space for next-generation sequencing studies of common genetic diseases. The familial evidence at each locus from one or more families is incorporated directly into the CLRT to increase the accuracy and greatly decrease the bioinformatic complexity of disease-gene identification efforts. We calculate statistical significance using a combination of permutation and gene-drop simulation to account for both the family structure and the observed pattern of variation in cases and controls. pVAAST supports dominant, recessive, and de novo inheritance models, and maintains high power across a wide variety of study designs, from monogenic, Mendelian diseases in a single family to highly polygenic, common diseases involving hundreds of families. We also demonstrate pVAASTs utility on exome chip, exome sequence, and whole-genome sequence data for recessive, dominant (cardiac septal defects), and complex genetic diseases (breast cancer, autism, and familial suicide). Our results demonstrate that pVAAST is a powerful and highly flexible tool for identifying disease genes in family-based sequencing studies.
You may contact the first author (during and after the meeting) at