Third-generation homozygosity analysis | LC26

Abstract
Runs of homozygosity (ROH) are extended chromosomal regions of consecutive homozygous genotypes arising from demographic history and recent parental relatedness through identity by descent (IBD), or from mitotic recombination events leading to uniparental disomy (UPD). Their identification is critical, as ROHs may harbor highly penetrant recessive disease-causing variants. Computationally, ROH detection relies on identifying consecutive homozygous single nucleotide polymorphisms (SNPs) and has traditionally been performed using SNP-array technologies. To meet the need for an analytical approach specifically designed for nanopore sequencing data, we developed a novel computational pipeline based on the H3M2 algorithm, capable of detecting ROHs from both low- and high-coverage nanopore datasets. The pipeline builds on the H3M2 algorithm, a hidden Markov model (HMM) modeling genomic distance between consecutive SNPs, enabling identification of ROHs across a broad range of genomic scales. We evaluated H3M2 using nanopore sequencing data generated at both high coverage (30x) and low coverage (2x) demonstrating that the method achieves ROH detection at a resolution comparable to that of SNP-array-based technologies. Notably, even when applied to low-coverage data, H3M2 identifies ROH regions with a resolution equivalent to that obtained using SNP arrays. Taken together, these results establish that our method fills a critical gap in nanopore sequencing by enabling the analysis of ROHs, a class of genomic events central to Mendelian disease genetics. Moreover, when combined with the ability of low-coverage nanopore sequencing to accurately identify high-resolution copy number variations, our findings indicate that low-coverage nanopore sequencing can fully replace current SNP- and comparative genomic hybridization (CGH)-array technologies within a single unified platform.

Biography
Alberto Magi is an Associate Professor of Bioengineering at the Department of Information Engineering, University of Florence. Since 2021, he has served as Scientific Director of the COMputational BIomediciNE Laboratory (COMBINE Lab), an interdisciplinary research unit involving four university departments. His research focuses on computational methods in biomedicine, with pioneering contributions to nanopore sequencing for liquid biopsy analysis, real-time detection of chromosomal anomalies in Mendelian diseases, and integrated genomic and epigenomic cancer profiling.