Fig. 1
From: Machine learning approaches to predict the Plant-associated phenotype of Xanthomonas strains
Workflow. Xanthomonas pathogenicity assay data for different strains was obtained from literature and stored in a SQL (phenotype) database. Available genomes were retrieved and de novo annotated with protein domains. Annotation results were stored in a Graph database. Strain specific domain content was used as input to train the classifiers. Resulting models were examined for their ability to predict pathogenicity and feature importance