A pipeline for the discovery of causal variants in Mendelian ME/CFS, 2026, Maccallini

[ME/CFS Science Blog]

A pipeline for the discovery of causal variants in Mendelian ME/CFS​

Paolo Maccallini

Identifying causal variants in suspected Mendelian disorders remains challenging when whole-genome sequencing reveals numerous rare candidates but no obvious molecular diagnosis. We developed PerGen, a pipeline that integrates family-based filtering of single-nucleotide variants, indels, structural variants, and mitochondrial variants with phenotype-driven prioritization, network analysis, and experimental validation. The pipeline integrates multiple existing tools (Seqr, Exomiser, AnnotSV, RegulomeDB, AMELIE, SpliceAI, CADD, DOMINO, dosage sensitivity metrics, and others) with custom R scripts. We applied the pipeline to whole-genome sequencing of a proband with early-onset, persistently housebound ME/CFS and three unaffected second-degree relatives.

To assess whether the proband represented an atypical form of ME/CFS, his phenotype was compared with that of 9,593 participants from the EMEA Pan-European ME/CFS Survey. The proband's Local Outlier Factor exceeded the 95th percentile of the reference distribution (p = 0.043), supporting investigation as a possible Mendelian case. After filtering, 19 candidate genes remained. Prioritization against a 369-gene ME/CFS disease module identified the neuronal adhesion genes CNTN6, CNTN5, and CDH13 as the strongest candidates (nominally significant), although none remained significant after correction for multiple testing. Nevertheless, these variants deserve further investigations in this case.

A previously unreported heterozygous splice-site variant in LBR (1:225411334 T>C on GRCh38) was absent from all sequenced relatives and predicted to abolish splice donor activity (SpliceAI = 0.73). Peripheral blood smear analysis demonstrated marked neutrophil hypolobulation (26% mononuclear, 68% bilobed, none with more than three lobes; n = 191), consistent with Pelger-Huët anomaly, supporting this variant as a novel pathogenic LBR allele. However, the benign nature of Pelger-Huët anomaly together with the weak network connectivity of LBR to the ME/CFS disease module suggests that this finding is incidental and unlikely to explain the patient's illness. Nevertheless, since compound heterozygous and homozygous mutations of LBR can lead to severe forms of skeletal dysplasia, this finding is relevant to genetic counseling.

Although no definitive molecular diagnosis was established, this study illustrates how integrated genomic analysis, phenotype outlier detection, and network-based prioritization can be combined to investigate suspected Mendelian forms of complex diseases.

Link to the paper: https://zenodo.org/records/21764136

Link to the blog introducing the paper: https://paolomaccallini.wordpress.c...overy-of-causal-variants-in-mendelian-me-cfs/

 

[ME/CFS Science Blog]

Much of this paper and the tools used are too complex for me. But it looks like an attempt at finding a rare mutation that might explain ME/CFS in a single patient.

Whole genome sequencing was performed by DanteLabs, and then various quality controls and filters were used to help identify rare mutations that might contribute to ME/CFS-symptoms. We each hold many rare mutations which are harmless so it's difficult to know if any of them is pathologic and related to ME/CFS. The paper links potential candidates with 369 genes that have been associated with ME/CFS in previous studies (115 from Synder's paper and 259 from PrecisionLife's analysis of DecodeME).

It highlights neuronal adhesion genes CNTN6, CNTN5, and CDH13 as the most likely candidates but the p-values didn't survive correction for multiple testing. I suppose it's also possible that there simply weren't any single mutations with large effects in relation to ME/CFS.

 

[ME/CFS Science Blog]

Two aunts and an uncle were also tested, presumable with the idea that rare mutations that also appear in family are unlikely to be damaging ones related to ME/CFS?