Biological Machine Learning Combined with Campylobacter Population Genomics Reveals Virulence Gene Allelic Variants Cause Disease

Abstract

Highly dimensional data generated from bacterial whole-genome sequencing is providing an unprecedented scale of information that requires an appropriate statistical analysis framework to infer biological function from populations of genomes. The application of genome-wide association study (GWAS) methods is an appropriate framework for bacterial population genome analysis that yields a list of candidate genes associated with a phenotype, but it provides an unranked measure of importance. Here, we validated a novel framework to define infection mechanism using the combination of GWAS, machine learning, and bacterial population genomics that ranked allelic variants that accurately identified disease. This approach parsed a dataset of 1.2 million single nucleotide polymorphisms (SNPs) and indels that resulted in an importance ranked list of associated alleles of porA in Campylobacter jejuni using spatiotemporal analysis over 30 years. We validated this approach using previously proven laboratory experimental alleles from an in vivo guinea pig abortion model. This framework, termed μPathML, defined intestinal and extraintestinal groups that have differential allelic porA variants that cause abortion. Divergent variants containing indels that defeated automated annotation were rescued using biological context and knowledge that resulted in defining rare, divergent variants that were maintained in the population over two continents and 30 years. This study defines the capability of machine learning coupled with GWAS and population genomics to simultaneously identify and rank alleles to define their role in infectious disease mechanisms.

Analyses of publicly available Hungatella hathewayi genomes revealed genetic distances indicating they belong to more than one species

“…In addition to taxonomic identification, genetic diversity, and virulence traits can be infer for disease potential. Additionally, allelic variants of core genes can be determined for disease presentation [15]. This is particularly useful in genomically diverse organisms with open pan-genomes and are common in the microbiome [15].…”

“…Additionally, allelic variants of core genes can be determined for disease presentation [15]. This is particularly useful in genomically diverse organisms with open pan-genomes and are common in the microbiome [15]. Previously, the genome sequence diversity of Hungatella was recognized to be larger than initially contemplated and led to the suggestion that the type strain is not representative of this species, as defined by the Human Microbiome Project [16].…”

Identification of Nitrogen Fixation Genes in Lactococcus Isolated from Maize Using Population Genomics and Machine Learning

“…For example, the resulting gene presence and absence information enables GWAS analysis in bacteria, where pangenome wide association studies (Pan-GWAS) with large populations of bacterial genomes serve as a method for the identification of genes associated with target phenotypes [ 30 ]. This Pan-GWAS approach was recently used to examine genetic deletions and successfully verified that only specific alleles of a core genome toxin gene ( porA ) caused abortion in livestock with Campylobacter jejuni infections with very high accuracy based on the population genomics approach [ 31 ]. While this strategy identifies genes of interest using widely accepted methods in statistical genetics, the implementation of machine learning algorithms to generate predictive models and identify genes with a ranked order of variable importance (VI) for targeted phenotypes is increasingly useful and provides an analytical method without a priori knowledge of the genes responsible for a trait [ 32 ].…”

Foodborne Pathogen Campylobacter

“…As with all modern science, the use of in silico bioinformatics approaches to study relevant research questions goes hand in hand with laboratory science. Bandoy and Weimer [19] use machine learning combined with Campylobacter population genomics to reveal virulence gene allelic variants that cause disease. The authors validated a novel framework to define infection mechanism using a combination of a GWAS, machine learning and bacterial population genomics that ranked allelic variants to identify disease.…”