Multi-omics analysis of pediatric diseases with Korean genome project

Abstract

In this Ph.D. dissertation, a multi-omics approach was applied to identify the characteristics and causes of pediatric diseases, and in particular, a large-scale Korean genome database was used in the studies to improve the accuracy of the analysis. The multi-omics approach used in this study provides a comprehensive understanding of various genomic and epigenomic variants, with a particular focus on whole-genome sequencing, whole-transcriptome sequencing, and methylation sequencing. This paper consists of three chapters, and the first chapter shows the suitability of using the Korean genome database for analysis. In the subsequent chapters, the findings and methodology from previous chapter were applied in the analysis of pediatric choroid plexus tumor (CPT) and rare disease patients to understand their mechanisms and identify their causative mutations. Chapter 1 introduces the Korea1K dataset as a valuable resource for the identification of potentially cancer-related mutations, providing an alternative to matched normal samples. Comparison with other global databases, such as 3.5KJPN and 1KGP, reveals the superiority of the Korea1K dataset in terms of accuracy and germline recovery when analyzing the genomes of Korean cancer patients. The study emphasizes the importance of ethnic group-specific genomic databases in cancer genomics research. The findings underscore the need for establishing large-scale genomic databases specific to ethnic groups to accurately characterize diseases and guide future research applications. Chapter 2 describes the comprehensive molecular characteristics of CPTs, with a particular focus on the difference between the CPT subtypes. A multi-omics analysis was performed on 20 pediatric CPT patients, encompassing choroid plexus papilloma (CPP), atypical choroid plexus papilloma (aCPP), mixed CPP and papillary ependymoma (mCPP), and choroid plexus carcinoma (CPC). Notably, CPCs specifically exhibited mutually exclusive TP53 and EPHA7 point mutations, along with the amplification of chromosome 1. Conversely, chromosome 9 amplification was specific to CPP. Genes associated with cell-cycle regulation and epithelial-mesenchymal transition pathways showed significant overexpression in CPC compared to CPP. Especially, CPC patients with leptomeningeal seeding (LMS) displayed notably elevated expression levels of genes associated with tumor metastasis and invasion compared to those without LMS. Additionally, the differential expression of AK1, regulated by both genomic and epigenomic factors, emerged as a contributing factor to the histological difference between CPP and CPC. Furthermore, methylation profiling revealed hypomethylation in major repeat regions, such as short interspersed nuclear elements, long interspersed nuclear elements, long terminal repeats, and retrotransposons, in CPC compared to CPP. These findings underscore the marked genetic and epigenetic distinctions between CPP and CPC. Chapter 3 consists of the analysis of trio-based whole-genome sequencing data of undiagnosed rare disease patients. There were around 50 patients who were undiagnosed even after panel or whole-exome sequencing. Causative variants were identified in approximately 50% of the undiagnosed patients. Among them, one patient was detected with a de novo variant in UBE2H as a candidate causative mutation. For functional validation of the detected variant, experiments were performed using zebrafish embryos. As a result, we confirmed that apoptosis is induced in UBE2H knockdown embryos. The analytical strategy developed in this study could be implemented to improve the diagnosis rate of Korean rare disease patients and provide individualized treatment strategies and monitoring for patients.