Seoul National University Biomedical Informatics (SNUBI), Division of Biomedical Informatics, Seoul National University College of Medicine, Seoul, 03080, Korea.
Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, M5G 2M9, Canada.
Pharmacogenomics. 2020 Aug;21(12):827-840. doi: 10.2217/pgs-2020-0039. Epub 2020 Jul 23.
Current gene-level prioritization methods aim to provide information for further prioritization of 'disease-causing' mutations. Since, they are inherently biased toward disease genes, methods specific to pharmacogenetic (PGx) genes are required. We proposed a gene-wise variant burden (GVB) method that integrates deleteriousness scores of the multitude of variants of a given gene at a personal-genome level. GVB in its simplest form outperformed the two state-of-the-art methods with regard to predicting pharmacogenes and complex disease genes but not for rare Mendelian disease genes. GVB* adjusted by paralog counts robustly performed well in most of the pharmacogenetic subcategories. Seven molecular genetic features well characterized the unique genomic properties of PGx, complex, and Mendelian disease genes. Altogether, GVB is an individual-specific genescore, especially advantageous for PGx studies.
目前的基因水平优先级方法旨在为进一步优先考虑“致病”突变提供信息。由于它们本质上偏向于疾病基因，因此需要特定于药物遗传学 (PGx) 基因的方法。我们提出了一种基于基因的变异负担 (GVB) 方法，该方法在个人基因组水平上整合了给定基因的多种变体的有害性评分。GVB 在最简单的形式下在预测药物基因和复杂疾病基因方面优于两种最先进的方法，但在罕见的孟德尔疾病基因方面并非如此。通过同源基因计数调整后的 GVB 在大多数药物遗传学亚类中表现稳健。七个分子遗传特征很好地描述了 PGx、复杂和孟德尔疾病基因的独特基因组特性。总的来说，GVB 是一种个体特异性基因评分，特别有利于 PGx 研究。
