Category: General | Date: Nov 26, 2025 | Author:
Our new open-source AI tool called Helixer can automatically identify genes in complex eukaryotic genomes using only DNA sequence, without needing extra data like RNA-seq. Developed by a team that has become international during the prohect and published in Nature Methods on 24 November 2025, Helixer combines deep learning with a hidden Markov model to produce gene annotations that approach curated reference quality across plants, animals, fungi and invertebrates. Nature
Its models generalize across species, allowing researchers to annotate newly sequenced genomes in a single command, greatly reducing the need for complex, data-hungry pipelines. Nature
We show that Helixer can help refine even well-studied genomes: in Arabidopsis thaliana, it filled gaps and corrected problematic gene models compared with existing reference annotations. Helixer is freely available for local installation via GitHub and can also be run through a web interface and the Galaxy ToolShed, making state-of-the-art genome annotation more accessible to the broader life-science community. Nature
General | Jan 14, 2026
By integrating deep single-cell RNA sequencing with spatial transcriptomics from...
General | Nov 26, 2025
Helixer a tool for structural annotation has been published in Nature Methods. T
General | Nov 19, 2025
We were hosting the nanopore day "The Power of Oxford Nanopore Technologies in Plant...
General | Nov 18, 2025
With our dear friends we managed to unravell the eggplant pangenome