Aim: A review of diagnostic outcomes from complementary massively parallel sequencing techniques in a large cohort of 213 families with neuromuscular disorders (NMD).
Methods: 85 trios and 128 individuals were subject to whole exome sequencing (WES) at the Broad Institute of Harvard and MIT or NMD panel screening at PathWest. 5/213 underwent additional whole genome sequencing (WGS; Broad Institute); 13/213 underwent additional muscle RNA sequencing (RNAseq; Broad Institute); 18/213 underwent both WGS and muscle RNAseq (Broad Institute). 15/213 had additional muscle mRNA laboratory investigations (Kids Neuroscience Centre).
Results: To date, a genetic diagnosis has been identified in 129/213 (61%) families. Six novel disease genes have been identified, with an average time of 31 months from identification of the gene, to E-publication in a peer reviewed journal. Seven novel or expanded phenotypes for known NMD genes were also determined. One third (42/129) of diagnosed families possessed at least one splicing variant.
WES alone provided a diagnosis for 92/129 (71%) diagnosed families. Additional sequencing and functional genomics pipelines (WGS, RNAseq, mRNA studies) were required for provision of a genetic diagnosis for 37/129 (29 %) diagnosed families. 29/37 (78%) families not diagnosed via initial WES screening, were shown subsequently to possess (at least one) splice-altering variant in known disease genes.
Discussion: Our results from a large cohort of families with NMD propose a step-wise approach to genomic analyses: initial screening via a targeted panel, triaging to WES, then WGS/RNA studies. Many cases solved by RNAseq and mRNA studies required additional insight from intronic variation provided by WGS. Functional genomics investigation to reach publication was lengthy (average 2.5 years), expensive and highly interdisciplinary.
Our collective data confirm that variants in common disease genes occur commonly – with ‘tricky variants’ in known disease genes (16%) more common than novel gene discovery (5%).