Integrating deep learning and unbiased automated high-content screening to identify complex disease signatures in human fibroblasts.

Drug discovery for diseases such as Parkinson's disease are impeded by the lack of screenable cellular phenotypes. We present an unbiased phenotypic profiling platform that combines automated cell culture, high-content imaging, Cell Painting, and deep learning. We applied this platform to primary fibroblasts from 91 Parkinson's disease patients and matched healthy controls, creating the largest publicly available Cell Painting image dataset to date at 48 terabytes. We use fixed weights from a convolutional deep neural network trained on ImageNet to generate deep embeddings from each image and train machine learning models to detect morphological disease phenotypes. Our platform's robustness and sensitivity allow the detection of individual-specific variation with high fidelity across batches and plate layouts. Lastly, our models confidently separate LRRK2 and sporadic Parkinson's disease lines from healthy controls (receiver operating characteristic area under curve 0.79 (0.08 standard deviation)), supporting the capacity of this platform for complex disease modeling and drug screening applications.

Diversity of the Trichocomaceae in the Katandra Nature Reserve, Central Coast, NSW, Australia.

The diversity of the family Trichocomaceae, which includes the major anamorph genera Aspergillus and Penicillium, was studied in the Katandra Nature Reserve, Central Coast, NSW, Australia. Soil, living leaves, leaf litter and detritus were examined by both direct and dilution plating techniques. Fungi were isolated on dichloran Rose Bengal chloramphenicol agar, and dichloran 18 % glycerol agar, media suitable for cultivation of many species within this family. Species of Trichocomaceae were isolated from all sites and all substrates examined. A high diversity was found, with more than 50 known species identified, and an equal number of undescribed species detected. More species of Penicillium were recovered than other genera, with Aspergillus species the next most common. Most of the species recovered were anamorphs, though 16 known and unknown ascosporic species were also isolated from heated and unheated soil. Soils, leaf litter, a scat from a native herbivore and leaves of living native plants yielded higher diversity than insects, worms or introduced plants. More species belonging to the family were isolated from soil in dry sclerophyll forest than in rainforest. Conversely, native rainforest plants harboured more diversity than the dry sclerophyll forest plants examined.

Phylogeography of the fungal pathogen Histoplasma capsulatum.

Until recently, Histoplasma capsulatum was believed to harbour three varieties, var. capsulatum (chiefly a New World human pathogen), var. duboisii (an African human pathogen) and var. farciminosum (an Old World horse pathogen), which varied in clinical manifestations and geographical distribution. We analysed the phylogenetic relationships of 137 individuals representing the three varieties from six continents using DNA sequence variation in four independent protein-coding genes. At least eight clades were idengified: (i) North American class 1 clade; (ii) North American class 2 clade; (iii) Latin American group A clade; (iv) Latin American group B clade; (v) Australian clade; (vi) Netherlands (Indonesian?) clade; (vii) Eurasian clade and (viii) African clade. Seven of eight clades represented genetically isolated groups that may be recognized as phylogenetic species. The sole exception was the Eurasian clade which originated from within the Latin American group A clade. The phylogenetic relationships among the clades made a star phylogeny. Histoplasma capsulatum var. capsulatum individuals were found in all eight clades. The African clade included all of the H. capsulatum var. duboisii individuals as well as individuals of the other two varieties. The 13 individuals of var. farciminosum were distributed among three phylogenetic species. These findings suggest that the three varieties of Histoplasma are phylogenetically meaningless. Instead we have to recognize the existence of genetically distinct geographical populations or phylogenetic species. Combining DNA substitution rates of protein-coding genes with the phylogeny suggests that the radiation of Histoplasma started between 3 and 13 million years ago in Latin America.