Analysis of Heterogeneous Genomic Samples Using Image Normalization and Machine Learning

Abstract

Background: Analysis of heterogeneous populations such as viral quasispecies is one of the most challenging bioinformatics problems. Although machine learning models are becoming to be widely employed for the analysis of sequencing data associated with such populations, their straightforward application is impeded by multiple challenges associated with technological limitations and biases, difficulty of selection of relevant features and need to compare genomic datastes of different sizes and structures. Methods:We propose a novel preprocessing approach to transform irregular genomic data into normalized image data. Such representation allows to restate the problems of classification and comparison of heterogeneous populations as image classification problems which can be solved using variety of available machine learning tools. We then apply the proposed approach to two important molecular epidemiology problems: inference of viral infection stage and detection of viral transmission clusters and outbreaks using next-generation sequencing data.Results: The infection staging method has been applied to HCV HVR1 samples collected from 108 recently and 257 chronically infected individuals. The SVM-based image classification approach achieved more than 95% accuracy for both recently and chronically HCV-infected individuals. Clustering has been performed on the data collected from 33 epidemiologically curated outbreaks, yielding more than 97% accuracy.Availability: The developed software is freely available at https://bitbucket.org/adv bio coll/chronic vs clinic

Epidemiological data analysis of viral quasispecies in the next-generation sequencing era

“…Measuring the time since infection using genomic data has recently been addressed in several studies [127][128][129][130][131]. The simpler version of this problem is infection staging, i.e.…”

“…distinguishing between recent and chronic infections using viral sequences sampled by NGS. A number of methods establish an age or stage of HIV or HCV infection using various measures of the population structure [127][128][129][130][131]. An underlying assumption of such methods is that intra-host viral evolution is associated with continuous genetic diversification.…”

“…To address this challenge, several filtering techniques have been proposed [149,150]. In consecutive studies [43,44,131,151], more sophisticated distance measures for quasispecies populations have been proposed. In particular, Melnyk et al [151] avoid reconstruction of haplotypes and/or phylogenetic trees by utilizing k-mer-based approach.…”