Facioscapulohumeral muscular dystrophy (FSHD) is the third-most common muscular dystrophy and affects hundreds of thousands of patients worldwide. Its pathogenesis involves complex genetic and epigenetic mechanisms at the D4Z4 locus, a macrosatellite array on chromosome 4q that normally comprises 8 to >100 3.3kb repeat units. Traditional methods of FSHD diagnosis rely on a series of separate assays for different markers of the disease: gel electrophoresis and Southern blotting for repeat number contractions (FSHD1) and detection of the pathogenic qA haplotype, sequencing for mutations in chromatin modifiers such as SMCHD1 (FSHD2), and bisulfite sequencing for D4Z4 hypomethylation (FSHD1&2). This process is slow, costly and resource-intensive, and only gives fragmented insights into the (epi)genetics of D4Z4 alleles.
Here we leverage ultra-long and Cas9-targeted long-read sequencing to develop a fast and accurate pipeline for comprehensively characterising the genetics and methylation of D4Z4 alleles. We use this to study samples from patients affected by FSHD1, FSHD2, and another disease caused by SMCHD1 variants, Bosma arhinia microphthalmia syndrome (BAMS), as well as publicly-available data from the 1000 Genomes Project. We attain high-read-depth sequencing of full-length D4Z4 arrays of up to 40 repeat units (132kb), accurately capture contracted arrays, genetic mosaicism, and pathogenic SMCHD1 variants, and generate accurate consensus sequences of D4Z4 alleles for variant analysis. Moreover, we identify and analyse complex D4Z4 structural variants including in-cis duplications, and reveal striking length- and SMCHD1 status-dependent methylation patterns across the D4Z4 array. Our findings provide new insights into human macrosatellite genetics and epigenetics, and demonstrate the potential of long-read Nanopore sequencing for improved FSHD diagnostics.