Classification and taxonomy articles from across Nature Portfolio

PepNet achieves accurate identification of both AIPs and AMPs by developing an interpretable neural network and applying a pre-trained protein language model.

Taxonomic classification of DNA sequences is typically performed on each sequence individually. Here the authors present a deep learning-based method that utilizes information across a dataset to enhance taxonomic annotations of any microbiome.

The Urfold model posits that different proteins may share architectural similarity despite topological variability. Here, an AI-based approach called DeepUrfold is developed to explore the Urfold idea at scale, revealing faint structural relationships across the protein universe.

Inspired by active learning approaches, we have developed a computational method that selects minimal gene sets capable of reliably identifying cell-types and transcriptional states in large sets of single-cell RNA-sequencing data. As the procedure focuses computational resources on poorly classified cells, active support vector machine (ActiveSVM) scales to data sets with over one million cells.

This month’s Genome Watch highlights the systematic discovery of defence systems, paving the way to decode novel genetic functions and further our understanding of microbial warfare.

A suite of new enzymes reveals more on how Nature breaks down plant-based polysaccharides and how these enzymes might be harnessed in the utilization of plant-based biomass.

Improvements to the Genome Taxonomy Database provide a complete bacterial and archaeal taxonomy.

Researchers identify a broad phylogenetic diversity of nucleocytoplasmic large DNA viruses by mining global-scale metagenomics datasets.

Complementary genomic frameworks for taxonomic classification of viruses infecting bacteria and archaea reveal evolutionary drivers, mosaicism and perspective on the genetic diversity of the tiniest, most abundant biological entities on Earth.