Reanalyzing clinical whole exome sequencing (WES) data aprovides additional diagnosis in paediatric-onset undiagnosed diseases – a Yale-HKU collaborative study

Abstract

Objective: To evaluate the effectiveness of WES reanalysis in undiagnosed patients in Hong Kong. Methods: This is a secondary analysis to Mak et al (2018) (PMID: 30109123), where 104 patients with suspected genetic disorder were recruited for WES in 2013-2017. The diagnostic yield was 40% (43/104). In this study, we included patients with no definitive diagnosis and 46 of them consented for reanalysis. Samples were either raw sequencing data obtained from clinical laboratories or DNA retrieved from HKU and sent to Yale for re-sequencing. Reads were aligned to hg19 reference genome using BWA and processed following GATK best practice guidelines including variant calling with HaplotypeCaller. Copy-number variants were called using gcnv (part of GATK4). Variants were uploaded to seqr for analysis. Results and Discussion: Among the 46 WES reanalyses (26 singletons, 1 duo and 19 trios), there were at least 11 definitive additional diagnoses. Eight are autosomal dominant conditions, with heterozygous pathogenic/likely pathogenic variants identified in PTPN11, PACS1, MFN2, SPTAN1 (2 patients), ATP1A3, GNB1 and MN1; three are autosomal recessive conditions (COQ7, SKIV2L and PRF1). By reviewing the 11 positive diagnoses, the majority (7/11; 64%) of patients harboured variants in genes that had weak/no gene-disease association in the initial analysis (PACS1, MFN2, SPTAN1, ATP1A3, GNB1 and COQ7), the pathogenicity became more confident with increasing literature. Four patients had a phenotype update or were presented with atypical symptoms that masked the core features (36%). Improved sequencing technology and bioinformatics also contributes to the increased yield (27%). Two variants had increased sequence coverage (PTPN11 and SKIV2L) and one exonic deletion was detected by new CNV caller (SPTAN1). The pathogenicity of MN1 was confirmed through international research collaboration. Conclusion: Reanalysis of WES in collaboration with a research centre provides promising additional diagnoses to patients, increasing the diagnostic yield from 40% to 52%. Increasing literature, evolving phenotype, improving analytical tools and international collaboration were the main reasons for positive findings in the reanalysis.