MADloy: Robust detection of mosaic loss of chromosome Y from genotype-array-intensity data

Abstract

Accurate protocols and methods to robustly detect the mosaic loss of chromosome Y (mLOY) are needed given its reported role in cancer, several age-related disorders and overall male mortality. Intensity SNP-array data have been used to infer mLOY status but there are discrepancies of reported findings due to the calling uncertainty of the methods and the differences in the tissue-matrix used. We proposed MADloy, a method that optimizes mLOY calling by correctly modeling the underlying reference population with no-mLOY status and incorporating B-deviation information. MADloy increased the calling accuracy with respect to previous methods in experimentally validated samples and improved the statistical power to detect associations with disease and mortality in simulation studies and real dataset analyses.Using MADloy in 18 individuals, we observed that mLOY cellularity slightly increased in blood after 3 years but was detectable in only 41% samples (7/17) in saliva. We found extreme down-regulation of the entire chromosome Y in kidney and bladder tumors with mLOY, while mLOY in blood was associated with deregulation of DNA replication, repair and recombination pathways. MADloy will help to define the mechanisms and roles of mLOY in different tissues as a male-specific marker of disease for personalized medicine.