Integrated Analysis of Genome and Transcriptome to Enhance Understanding of Rare Neuromuscular Disorders

whole exome sequencing;transcriptome sequencing;neuromuscular disorders;multiomics;diagnosis;Mendelian disorder;variant discovery;expression-based clustering;610.72

Jana KneisslNext generation sequencing ;

University:서울대학교 대학원

关键词: whole exome sequencing;    transcriptome sequencing;    neuromuscular disorders;    multiomics;    diagnosis;    Mendelian disorder;    variant discovery;    expression-based clustering;    610.72;   

Others  :  http://s-space.snu.ac.kr/bitstream/10371/161542/1/000000157454.pdf

美国|英语

来源: Seoul National University Open Repository

【 摘 要 】

Introduction. Whole exome sequencing has become a robust and standard tool for rare diseases diagnosis thanks to advantages in cost and data handling. However, whole exome sequencing-based diagnosis rates typically do not exceed 50%, which can be attributed to the difficulty of interpreting variants of uncertain significance, as well as to the disregard of non-coding variants, including variants in intronic and regulatory regions in the genome. Therefore, I explored the utility of transcriptome sequencing as a compensatory approach in rare neuromuscular disorders diagnosis.Methods. Whole exome sequencing of 94 patients with undiagnosed neuromuscular disorders was collected from Seoul National University Children’s Hospital and analyzed for variants in known neuromuscular disease genes. Additional transcriptome sequencing was performed for 63 of the whole exome sequenced patients and for ten patients without genome data. Transcriptome data were utilized for cryptic damaging variants, differentially expression, aberrant splicing and allele specific expression analysis. Furthermore, non-negative matrix factorization was applied to identify expression-based clustering and cluster-specific gene ontology was derived.Results. Whole exome sequencing analysis identified candidate variants in 49% of patients, with 83% of them located within known disease genes. Structural variants with questionable pathogenicity were discovered in twelve cases. RNA-Sequencing based variant calling lead to further discovery of heterozygous candidate variants in nine samples, five of which did not undergo whole exome sequencing. Allele specific expression identified two likely candidate genes and differential gene expression analysis lead to the prioritization of sets of genes in an additional four samples. Lastly, aberrant splicing of DMD, TTN and MICU1 was detected in each of four samples. Non-negative matrix factorization-based clustering resulted in the identification of six clusters with distinct gene expression profiles.Discussion. Firstly, I aimed to evaluate whether transcriptome sequencing can provide additional evidence for the interpretation of whole exome sequencing variants. Overall, transcriptome sequencing was able to detect abnormalities associated with the previously identified mutation in less than 30% of positive whole exome sequencing cases. For samples without whole exome sequencing result, I successfully used transcriptome sequencing to identify potential pathogenic causes in 18 cases. In conclusion, transcriptome sequencing proved to be a useful tool for the diagnosis of whole exome sequencing negative samples, but did not prove to have great utility for the interpretation of pathogenic whole exome sequencing variants.