phenotypic heterogeneity is an important trait for the development and survival of many microorganisms including the yeast Cryptococcus spp., a deadly pathogen spread worldwide. Here, we have applied scanning electron microscopy (SEM) to define four Cryptococcus spp. capsule morphotypes, namely Regular, Spiky, Bald, and Phantom. These morphotypes were persistently observed in varying proportions among yeast isolates. to assess the distribution of such morphotypes we implemented an automated pipeline capable of (1) identifying potentially cell-associated objects in the SEM-derived images; (2) computing object-level features; and (3) classifying these objects into their corresponding classes. The machine learning approach used a Random Forest (RF) classifier whose overall accuracy reached 85% on the test dataset, with per-class specificity above 90%, and sensitivity between 66 and 94%. Additionally, the RF model indicates that structural and texture features, e.g., object area, eccentricity, and contrast, are most relevant for classification. The RF results agree with the observed variation in these features, consistently also with visual inspection of SEM images. Finally, our work introduces morphological variants of Cryptococcus spp. capsule. these can be promptly identified and characterized using computational models so that future work may unveil morphological associations with yeast virulence. open Scientific RepoRtS | (2020) 10:2362 | https://doi.org/10.1038/s41598-020-59276-w www.nature.com/scientificreports www.nature.com/scientificreports/ yeast kill 180,000 people around the world every year 13 . In addition, the treatment of cryptococcosis is unaffordable in most in developing countries 14,15 .An important trait for cryptococcal virulence is the presence of a polysaccharide capsule, which has been well described at the molecular and functional levels 16,17 . Capsular morphology includes a huge heterogeneity among distinct isolates and even in clonal isolates. This cellular diversity has been exploited in terms of drug resistance and pathogenic potential [18][19][20] .The study of phenotype heterogeneity is hampered by the lack of proper detection methods and statistical analyses. In order to further explore this morphotype diversity, we implemented an automated image analysis pipeline. This machine learning approach is capable of detecting and classifying capsular morphotypes, being applicable to cell type quantification in microscopy-based experiments. While not a high-throughput technique per se, SEM does yield vast amounts of complex data. Here, we describe the adaptation of one algorithmic implementation for the analysis and classification of Cryptococcus spp. capsular morphotypes captured using scanning electron microscopy (SEM). Our model substantially increases data analysis efficiency and provides a template for future machine learning applications within microbiology.
ResultsCryptococcus spp. exhibit distinct capsule morphotypes under scanning electron microscopy. The analysis of Cryptococcus s...