Tiberius: end-to-end deep learning with an HMM for gene prediction

2024/11/13 by Lars Gabriel, Felix Becker, Katharina J. Hoff +1 · 2 voices · 38 citations
Biochemistry, Genetics and Molecular Biology · #Ab initio #Artificial intelligence #Biology #Computational biology #Computer science #Deep learning #Gene #Gene prediction #Genetics #Genome #Genomics and Chromatin Dynamics #Genomics and Phylogenetic Studies #Hidden Markov model #Human genome #Machine Learning in Bioinformatics #Machine learning #Physics

paper · doi:10.1093/bioinformatics/btae685

published in Bioinformatics 40(12) (Oxford University Press)

openalex publication_date 2024/11/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

MOTIVATION: For more than 25 years, learning-based eukaryotic gene predictors were driven by hidden Markov models (HMMs), which were directly inputted a DNA sequence. Recently, Holst et al. demonstrated with their program Helixer that the accuracy of ab initio eukaryotic gene prediction can be improved by combining deep learning layers with a separate HMM postprocessor. RESULTS: We present Tiberius, a novel deep learning-based ab initio gene predictor that end-to-end integrates convolutional and long short-term memory layers with a differentiable HMM layer. Tiberius uses a custom gene prediction loss and was trained for prediction in mammalian genomes and evaluated on human and two other genomes. It significantly outperforms existing ab initio methods, achieving F1 scores of 62% at gene level for the human genome, compared to 21% for the next best ab initio method. In de novo mode, Tiberius predicts the exon-intron structure of two out of three human genes without error. Remarkably, even Tiberius's ab initio accuracy matches that of BRAKER3, which uses RNA-seq data and a protein database. Tiberius's highly parallelized model is the fastest state-of-the-art gene prediction method, processing the human genome in under 2 hours. AVAILABILITY AND IMPLEMENTATION: https://github.com/Gaius-Augustus/Tiberius.

Cited by

• Cookbook for plant genome sequences
• Annotation matters: the effect of structural gene annotation on orthology inference
• Improving spliced alignment by modeling splice sites with deep learning
• Accurate ab initio gene prediction in eukaryotes with Tiberius in multiple clades
• Annotation of 200 Insect Genomes with BRAKER for Consistent Comparisons across Species
• Chimeric mis-annotations of genes remain pervasive in eukaryotic non-model organisms
• GeneCAD: Plant Genome Annotation with a DNA Foundation Model
• Helixer: ab initio prediction of primary eukaryotic gene models combining deep learning and a hidden Markov model
• Ragnarok: a flexible and RApid GeNe Annotation (ROcKs) pipeline deployed through Nextflow
• A chromosomal level genome assembly of Nguni Sheep, Ovis aries
• Chimeric mis-annotations of genes remain pervasive in eukaryotic non-model organisms. [europepmc]
• NetStart 2.0: prediction of eukaryotic translation initiation sites using a protein language model. [europepmc]
• Helixer: ab initio prediction of primary eukaryotic gene models combining deep learning and a hidden Markov model. [europepmc]
• Ensembl 2026. [europepmc]
• An overview of computational methods for gene prediction in eukaryotes: strengths, limitations, and future directions. [europepmc]
• Improving spliced alignment by modeling splice sites with deep learning. [europepmc]
• A high-quality chromosome level genome assembly of the South African indigenous Nguni goat (Capra hircus). [europepmc]
• Cookbook for plant genome sequences. [europepmc]
• Improved reconstruction of transcripts and coding sequences from RNA-seq data. [europepmc]
• Annotation of 200 Insect Genomes with BRAKER for Consistent Comparisons across Species. [europepmc]
• Multi-feature fusion for gene prediction and functional peptide identification. [europepmc]
• A chromosome-level genome assembly of the South African indigenous, Kolbroek pig, Sus scrofa domesticus. [europepmc]
• Artificial Intelligence in genomics: a comprehensive survey of methods, resources, challenges, and prospects. [europepmc]

Discussions

• Big news! Tiberius, our new deep learning tool for genome annotation has been accepted for publication at Bioinformatics :-) doi.org/10.1093/bioi... [bsky, 25 points, 3 comments]
• Tiberius: End-to-end deep learning with an HMM for gene prediction doi.org/10.1093/bioi... #bioinformatics #AI #deeplearning [bsky, 2 points, 0 comments]

Related

• Tiberius: End-to-End Deep Learning with an HMM for Gene Prediction
• Accurate ab initio gene prediction in eukaryotes with Tiberius in multiple clades
• Helixer: ab initio prediction of primary eukaryotic gene models combining deep learning and a hidden Markov model
• DeepMetabolism: A Deep Learning System to Predict Phenotype from Genome Sequencing
• Enigma: An Efficient Model for Deciphering Regulatory Genomics