Improved Gene Annotation of the Fungal Wheat Pathogen <i>Zymoseptoria tritici</i> Based on Combined Iso-Seq and RNA-Seq Evidence
Nicolas Lapalu, Lucie Lamothe, Yohann Petit, Anne Génissel, Camille Delude, Alice Feurtey, Leen Abraham, Daniel P. Smith, Robert C. King, Alison Renwick, Mélanie Appertet, Justine Sucher, Andrei Stecca Steindorff, Stephen B. Goodwin, G.H.J. Kema, Igor V. Grigoriev, James K. Hane, J. J. Rudd, Eva Stukenbrock, Daniel Croll, Gabriel Scalliet, Marc‐Henri Lebrun
Abstract
Despite large omics datasets, the prediction of eukaryotic genes is still challenging. We have developed a new method to improve the prediction of eukaryotic genes and demonstrate its utility using the genome of the fungal wheat pathogen Zymoseptoria tritici. From 10,933 to 13,260 genes were predicted by four previous annotations, but only one third were identical. A novel bioinformatics suite, InGenAnnot, was developed to improve Z. tritici gene annotation using Iso-Seq full-length transcript sequences. The best gene models were selected among different ab initio gene predictions, according to transcript and protein evidence. Overall, 13,414 reannotated gene models (RGMs) were predicted, improving previous annotations. Iso-Seq transcripts outlined 5′ and 3′ untranslated regions for 73% of the RGMs and alternative transcripts mainly due to intron retention. Our results showed that the combination of different ab initio gene predictions and evidence-driven curation improved gene annotation of a eukaryotic genome. It also provided new insights into the transcriptional landscape of this fungus. [Formula: see text] Copyright © 2025 The Author(s). This is an open access article distributed under the CC BY 4.0 International license .