137. Estimation of familial DNA contamination using Mendelian inconsistencies from next generation sequencing of trios

Abstract

Christopher Yoon^a, Malachi Griffith^a, Obi L. Griffith^a, Young Seok Ju^b

^aWashington University, St. Louis, MO, USA; ^bKorea Advanced Institute of Science and Technology, Daejeon, Daejeon, Korea

Cancer is often hereditary which would require sequencing of the proband as well as the parents. Before detecting and analyzing cancer mutations, quality control of the family genome sequencing is needed to ensure integrity of the data. However, accurate detection and quantification of the contamination in a DNA mixture is particularly difficult when genetically related family members are the sources. We developed a maximum likelihood estimation framework based on Mendel’s law of inheritance, to quantify the degree of mixed DNA between family members in whole genome sequencing data of parent-offspring trios. We accurately quantified intrafamilial DNA contamination including parent-offspring, sibling-sibling and multiple familial sources. In addition, we identified uniparental disomy and chimerism that violate Mendelian inheritance patterns. These atypical genomes would need special considerations when conducting family studies, including the study of hereditary cancer. We tested our tool on both simulated and real data sets to show it can detect various intrafamilial contaminations and atypical genomes in family sequencing studies.