Variant Discovery Using Next-Generation Sequencing and its Future Role in Pharmacogenetics

Abstract

Next-generation sequencing (NGS) has enabled the discovery of a multitude of novel and mostly rare variants in pharmacogenes that may alter a patient’s therapeutic response to drugs. In addition to single nucleotide variants, structural variation affecting the number of copies of whole genes or parts of genes can be detected. While current guidelines concerning clinical implementation mostly act upon well-documented, common single nucleotide variants to guide dosing or drug selection, in silico and large-scale functional assessment of rare variant effects on protein function are at the forefront of pharmacogenetic research to facilitate their clinical integration. Here, we discuss the role of NGS in variant discovery, paving the way for more comprehensive genotype-guided pharmacotherapy that can translate to improved clinical care.

Potential of Whole-Genome Sequencing-Based Pharmacogenetic Profiling

“…Although SRS is powerful, it comes with inherent limitations. In contrast, long-read sequencing (LRS; read lengths >1 kilobases) for example, by Pacific Biosciences (pacb.com; PacBio) and Oxford Nanopore Technologies (nanoporetech.com; ONT) has the potential to overcome these limitations, but has often been dismissed for being too expensive and having raw-read error rates of ∼10%, compared with <1% using SRS [36,57]. However, along with other recent seminal studies [23,58,59], the recently published precisionFDA Truth Challenge V2, which aimed to call sequence variants in difficult-to-map regions, emphasized that LRS alone or in combination with SRS is the superior tool for variant calling, phasing and resolving complex or repetitive genomic regions, including several pharmacogenes [60].…”

“…However, the most efforts of variant effect prediction have focused on the identification of damaging/deleterious variants, in other words, the distinction from neutral variants. In silico algorithms are either based on physiochemical differences among amino acids (e.g., Grantham score [80]), phylogenetic conservation (e.g., GERP, SiPhy) and/or effect on protein structure (e.g., Polyphen-2, SIFT) or ensemble tools combining various existing tools into a single score (e.g., CADD, DANN, Eigen, REVEL) [57]. These algorithms, however, are not optimized for often evolutionary diverse pharmacogenes.…”

Pharmacogenomics in the era of next generation sequencing – from byte to bedside

“…Importantly however, SRS faces major limitations for the profiling of complex or repetitive genetic loci, as short reads are often difficult to unambiguously align A c c e p t e d M a n u s c r i p t or assemble, resulting in issues with the detection of large structural variation and variant phasing 67,68 . These shortcomings are particularly relevant in pharmacogenomics, as many relevant genes, such as CYP2D6, CYP2A6, ABCB1, SLC22A1 and HLA genes, are highly polymorphic with nearby considerable intervals of low complexity regions, as well as segmental duplications or variable number tandem repeats 69 .…”

“…A variety of cellular phenotypes can be identified using this approach, including protein abundance and binding or metabolism of substrates for variants within However, while such highly multiplexed strategies are becoming increasingly adopted, limitations remain, including the necessity to develop and optimize specific selection and screening assays for each gene of interest, and limited expertise in using these approaches 67,104 . Furthermore, these assays cannot assess certain important aspects related to pharmacogenetic variants, including potential implications of post-translational modifications or cellular localization of the protein of interest.…”

Machine Learning and Deep Learning for the Pharmacogenomics of Antidepressant Treatments

“…On another note, the typical concerns in deep learning, namely deep neural networks or deep MFNNs, may be included in the aforementioned studies [ 63 ]. For example, the accuracy of the deep learning frameworks would be relatively flat when the cohort size is low [ 68 ]. Moreover, it is pivotal that a number of independent experiments would have generalized their findings by using universal benchmark datasets [ 69 ].…”