In PLOS Neglected Tropical Diseases, a team from Sudan, the Netherlands, and elsewhere presents findings from a molecular study of Madurella fungal species behind a type of subcutaneous infection called black-grain eumycetoma in central Sudan. When they assessed mycetoma biopsy samples from more than 100 individuals treated for infections at foot, hand, or other sites at a center in the state of Gezira with the help of PCR amplification and targeted sequencing, the researchers identified the M. mycetomatis species in around 88 percent of confirmed black-grain eumycetoma cases, though examples of infections involving M. fahalii, M. tropicana, and other species also turned up. "The identification of uncommon agents of eumycetoma in this study, such as M. fahalii, M. tropicana, and [Sphaerulina rhododendricola] highlighted the importance of species identification in clinical settings," they write, adding that "[t]hese species might differ in their susceptibility to antifungal agents and their epidemiology."
Australian researchers report on results from a food-focused analysis of Campylobacter species for a study appearing in PLOS One. Using whole-genome sequencing, the team took a look at hundreds of C. coli and C. jejuni isolates from more than 600 raw meat samples sold at retail stories in several Australian states, highlighting some 113 new and known Campylobacter sequence types in the chicken, lamb, beef, and pork samples tested. From these and other results, the authors conclude that "Campylobacter species from retail products in Australia are highly genotypically diverse and important differences in antimicrobial resistance exist between Campylobacter species and animal sources."
In another PLOS One paper, a University of Strasbourg team introduces an algorithm aimed at sifting through variants of unknown significance to find missense variants involved in disease. The prediction tool — dubbed "missense deleteriousness predictor," or MISTIC — brings together features from existing machine learning algorithms to produce a supervised machine-learning model for finding deleterious missense changes in exome sequence data, the researchers say. After training MISTIC with a range of features linked to more than 100 known missense mutations, the authors found that the prediction tool compared favorably to those used for classifying variants in the past, though they note that "[f]uture improvements will include additional informative features, such as multi-ethnic [minor allele frequencies] from other population databases, genotype frequencies, and gene-based calibration of the different scores."