DL-ADR: a novel deep learning model for classifying genomic variants into adverse drug reactions

Abstract

BackgroundGenomic variations are associated with the metabolism and the occurrence of adverse reactions of many therapeutic agents. The polymorphisms on over 2000 locations of cytochrome P450 enzymes (CYP) due to many factors such as ethnicity, mutations, and inheritance attribute to the diversity of response and side effects of various drugs. The associations of the single nucleotide polymorphisms (SNPs), the internal pharmacokinetic patterns and the vulnerability of specific adverse reactions become one of the research interests of pharmacogenomics. The conventional genomewide association studies (GWAS) mainly focuses on the relation of single or multiple SNPs to a specific risk factors which are a one-to-many relation. However, there are no robust methods to establish a many-to-many network which can combine the direct and indirect associations between multiple SNPs and a serial of events (e.g. adverse reactions, metabolic patterns, prognostic factors etc.). In this paper, we present a novel deep learning model based on generative stochastic networks and hidden Markov chain to classify the observed samples with SNPs on five loci of two genes (CYP2D6 and CYP1A2) respectively to the vulnerable population of 14 types of adverse reactions.MethodsA supervised deep learning model is proposed in this study. The revised generative stochastic networks (GSN) model with transited by the hidden Markov chain is used. The data of the training set are collected from clinical observation. The training set is composed of 83 observations of blood samples with the genotypes respectively on CYP2D6*2, *10, *14 and CYP1A2*1C, *1 F. The samples are genotyped by the polymerase chain reaction (PCR) method. A hidden Markov chain is used as the transition operator to simulate the probabilistic distribution. The model can perform learning at lower cost compared to the conventional maximal likelihood method because the transition distribution is conditional on the previous state of the hidden Markov chain. A least square loss (LASSO) algorithm and a k-Nearest Neighbors (kNN) algorithm are used as the baselines for comparison and to evaluate the performance of our proposed deep learning model.ResultsThere are 53 adverse reactions reported during the observation. They are assigned to 14 categories. In the comparison of classification accuracy, the deep learning model shows superiority over the LASSO and kNN model with a rate over 80 %. In the comparison of reliability, the deep learning model shows the best stability among the three models.ConclusionsMachine learning provides a new method to explore the complex associations among genomic variations and multiple events in pharmacogenomics studies. The new deep learning algorithm is capable of classifying various SNPs to the corresponding adverse reactions. We expect that as more genomic variations are added as features and more observations are made, the deep learning model can improve its performance and can act as a black-box but reliable verifier for other GWAS studies.

Deep Learning in Pharmacogenomics: From Gene Regulation to Patient Stratification

“…Applications of deep learning recently demonstrate state-of-the-art performance for predicting cell phenotypes from transcriptomics data [122], drug response in cancer [123], seizure-inducing side effects of preclinical drugs [124], patient survival from multi-omic data [38], drug-induced liver injury prediction [62], and classifying genomic variants into adverse drug reactions [125].…”

Phenotype Prediction and Genome-Wide Association Study Using Deep Convolutional Neural Network of Soybean

“…Recently, deep learning has been applied in computational biology (Angermueller et al, 2016), with the introduction of noncoding variant function prediction (Zhou and Troyanskaya, 2015), protein localization prediction (Alipanahi et al, 2015; Zhang N et al, 2018), protein secondary structure prediction (Spencer et al, 2015), and protein post-translational modification site prediction (Wang D et al, 2017; Wang et al, 2018). In genotype association studies, deep learning has also been used to identify SNP interactions (Uppu et al, 2016), classify genomic variants (Liang et al, 2016). DeepGS, an ensemble of convolutional neural network (CNN) (Krizhevsky et al, 2012) and rrBLUP have been used to predict phenotypes using imputed SNPs (Ma et al, 2018), and a simple dense neural network (DNN) is used on genotype-by-sequencing (GBS) data (Montesinos-López et al, 2018).…”

Deep learning: from chemoinformatics to precision medicine

“…Population variability can provide insight into drug exposure and response variations within a population (Cubitt et al 2011;Jamei et al 2009). Recently, a Markov chain process was used to classify samples with single nucleotide polymorphisms on five loci of two drug metabolizing enzyme genes CYP2D6 and CYP1A2 in a vulnerable population with 14 types of ADR including platelet count, hemoglobin, and urine protein abnormalities (Liang et al 2016). A deep learning-based algorithmic framework, DeepSEA was developed to predict noncoding-variant effects de novo from sequence (Zhou and Troyanskaya 2015), which can help to identify noncoding genomic regions and the potential functions of polymorphisms associated with complex diseases or traits.…”