D4Z4 Methylation Levels Combined with a Machine Learning Pipeline Highlight Single CpG Sites as Discriminating Biomarkers for FSHD Patients

Abstract

The study describes a protocol for methylation analysis integrated with Machine Learning (ML) algorithms developed to classify Facio-Scapulo-Humeral Dystrophy (FSHD) subjects. The DNA methylation levels of two D4Z4 regions (DR1 and DUX4-PAS) were assessed by an in-house protocol based on bisulfite sequencing and capillary electrophoresis, followed by statistical and ML analyses. The study involved two independent cohorts, namely a training group of 133 patients with clinical signs of FSHD and 150 healthy controls (CTRL) and a testing set of 27 FSHD patients and 25 CTRL. As expected, FSHD patients showed significantly reduced methylation levels compared to CTRL. We utilized single CpG sites to develop a ML pipeline able to discriminate FSHD subjects. The model identified four CpGs sites as the most relevant for the discrimination of FSHD subjects and showed high metrics values (accuracy: 0.94, sensitivity: 0.93, specificity: 0.96). Two additional models were developed to differentiate patients with lower D4Z4 size and patients who might carry pathogenic variants in FSHD genes, respectively. Overall, the present model enables an accurate classification of FSHD patients, providing additional evidence for DNA methylation as a powerful disease biomarker that could be employed for prioritizing subjects to be tested for FSHD.

Association of 4qA-Specific Distal D4Z4 Hypomethylation With Disease Severity and Progression in Facioscapulohumeral Muscular Dystrophy

“…Similar issues with inconclusive diagnostic analysis of FSHD based on detection of 4qA methylation levels have also been noted in recent studies. 21,22 One recent study proposed that D4Z4 hypomethylation could serve as a potentially important prognostic marker for FSHD 21 but contradicted another report that found no correlation between D4Z4 methylation levels assessed by MSRE and clinical manifestation or disease severity. 23 In this study, we stratified patients according to CpG6 methylation for univariate analysis of disease severity, which indicated that muscle weakness (measured by CS and ACSS) is more severe, and onset age of first-ever muscle weakness is earlier in patient groups with lower methylation levels (groups of LM1, LM2, and LM3).…”

“…Similar issues with inconclusive diagnostic analysis of FSHD based on detection of 4qA methylation levels have also been noted in recent studies. 21,22…”

The FSHD jigsaw: are we placing the tiles in the right position?

“…Recently, Caputo and colleagues [57,58] developed a protocol for methylation analysis of specific CpG residues using iMachine Learning (ML) algorithms to classify FSHD cases. Hiramuki et al , [59 ▪ ] applied long-read sequencing through a Nanopore CRISPR/Cas9-targeted resequencing to diagnose FSHD by simultaneous detection of D4Z4 repeat length and methylation status at the nucleotide level in FSHD patients.…”

Epigenetic profiling of the D4Z4 locus: Optimization of the protocol for studying DNA methylation at single CpG site level

“…The study involved 335 subjects (160 FSHD patients and 175 healthy subjects), which were available from a previous study [23]. In particular, the analysis of both cohorts allowed testing the ability of the assay to detect all the methylated strands independently from the disease condition.…”

“…In particular, the BSS technique allows detecting methylated and unmethylated cytosines (C) with a single‐base resolution, which, in turn, enables the assessment of methylation levels related to single CpG sites. This method has been recently applied to evaluate the methylation levels of two regions (viz., the DUX4 ‐PAS and the DR1) in FSHD patients [23].…”