Software articles from across Nature Portfolio

StringTie3 shows enhanced performance in total RNA-seq assembly and quantification by modeling both nascent and mature transcripts, across short-, long- and hybrid-read sequencing data.

Spatial transcriptomics platforms span multicellular to subcellular resolutions but lack methods to reconstruct single-cell transcriptomes across them. Here, the authors introduce STARS, a unified deep learning framework that integrates histology and transcriptomics to infer single-cell gene expression across resolutions.

A method involving class-associated manifold learning allows the logic underlying black-box medical AI models to be intuitively explained to accurately clarify the behaviour of medical AI models and improve medical knowledge discovery.

Model building in cryo-EM strongly depends on its map quality. Here, the authors propose EMReady2, a generalist tool for fast improvement of cryo-EM and cryo-ET maps at 2–10 Å resolutions with local quality-aware deep learning.

Metagenomic analyses of microbial communities can be compromised by technical biases including cross-sample contamination. Here, the authors present CroCoDeEL, a control-free method for detecting cross-sample contamination in metagenomics.

Recent price hikes, usage limitations and unreliable outputs are causing some scientific researchers to think twice about using artificial intelligence.

AI foundation models (FMs) are transforming cancer management and research. Here, we analyze the state-of-the-art FMs, discuss their impact on cancer management, and argue that the next generation of cancer AI FMs will be defined by multimodality, enhanced reasoning, maximized openness, and sustained human guidance.

The open-source Orthrus RNA language model maps evolutionary patterns across mammalian species to predict key mRNA properties, advancing RNA biology research.

We introduce neural fields for adaptive optical two-photon fluorescence microscopy (NeAT), a framework that reconstructs 3D structure from a volumetric image stack while estimating wavefront aberrations and correcting sample motion and conjugation errors. Application of NeAT on custom and commercial microscopes improved quality of in vivo structural and functional imaging of the mouse brain.