With the high prevalence of breast cancer, it is urgent to find out the intrinsic difference between various subtypes, so as to infer the underlying mechanisms. Given the available multi-omics data, their proper integration can improve the accuracy of breast cancer subtype recognition. In this study, DeepMO, a model using deep neural networks based on multi-omics data, was employed for classifying breast cancer subtypes. Three types of omics data including mRNA data, DNA methylation data, and copy number variation (CNV) data were collected from The Cancer Genome Atlas (TCGA). After data preprocessing and feature selection, each type of omics data was input into the deep neural network, which consists of an encoding subnetwork and a classification subnetwork. The results of DeepMO based on multi-omics on binary classification are better than other methods in terms of accuracy and area under the curve (AUC). Moreover, compared with other methods using single omics data and multi-omics data, DeepMO also had a higher prediction accuracy on multi-classification. We also validated the effect of feature selection on DeepMO. Finally, we analyzed the enrichment gene ontology (GO) terms and biological pathways of these significant genes, which were discovered during the feature selection process. We believe that the proposed model is useful for multi-omics data analysis.

Expansion of penguin activity in maritime Antarctica, under ice thaw, increases the chances of penguin feces affecting soil microbiomes. The detail of such effects begins to be revealed. By comparing soil geochemistry and microbiome composition inside (one site) and outside (three sites) of the rookery, we found significant effects of penguin feces on both. First, penguin feces change soil geochemistry, causing increased moisture content (MC) of ornithogenic soils and nutrients C, N, P, and Si in the rookery compared to non-rookery sites, but not pH. Second, penguin feces directly affect microbiome composition in the rookery, not those outside. Specifically, we found 4,364 operational taxonomical units (OTUs) in 404 genera in six main phyla: Proteobacteria, Actinobacteria, Gemmatimonadetes, Acidobacteria, Chloroflexi, and Bacteroidetes. Although the diversity is similar among the four sites, the composition is different. For example, penguin rookery has a lower abundance of Acidobacteria, Chloroflexi, and Nitrospirae but a higher abundance of Bacteroidetes, Firmicutes, and Thermomicrobia. Strikingly, the family Clostridiaceae of Firmicutes of penguin-feces origin is most abundant in the rookery than non-rookery sites with two most abundant genera, Tissierella and Proteiniclasticum. Redundancy analysis showed all measured geochemical factors are significant in structuring microbiomes, with MC showing the highest correlation. We further extracted 21 subnetworks of microbes which contain 4,318 of the 4,364 OTUs using network analysis and are closely correlated with all geochemical factors except pH. Our finding f penguin feces, directly and indirectly, affects soil microbiome suggests an important role of penguins in soil geochemistry and microbiome structure of maritime Antarctica.

We pragmatically delineate the trait repertoire that enables organismal niche specialization. We based our approach on the tenet, derived from evolutionary and complex-system considerations, that genomic units that can significantly contribute to fitness in a certain habitat will be comparatively more complex in organisms specialized to that habitat than their genomic homologs found in organisms from other habitats. We tested this in cyanobacteria forming harmful water blooms, for which decades-long efforts in ecological physiology and genomics exist. Our results essentially confirm that genomics and ecology can be linked through comparative complexity analyses, providing a tool that should be of general applicability for any group of organisms and any habitat, and enabling the posing of grounded hypotheses regarding the ecogenomic basis for diversification.