Helixer: cross-species gene annotation of large eukaryotic genomes using deep learning

Abstract

Motivation Current state-of-the-art tools for the de novo annotation of genes in eukaryotic genomes have to be specifically fitted for each species and still often produce annotations that can be improved much further. The fundamental algorithmic architecture for these tools has remained largely unchanged for about two decades, limiting learning capabilities. Here, we set out to improve the cross-species annotation of genes from DNA sequence alone with the help of deep learning. The goal is to eliminate the dependency on a closely related gene model while also improving the predictive quality in general with a fundamentally new architecture. Results We present Helixer, a framework for the development and usage of a cross-species deep learning model that improves significantly on performance and generalizability when compared to more traditional methods. We evaluate our approach by building a single vertebrate model for the base-wise annotation of 186 animal genomes and a separate land plant model for 51 plant genomes. Our predictions are shown to be much less sensitive to the length of the genome than those of a current state-of-the-art tool. We also present two novel post-processing techniques that each worked to further strengthen our annotations and show in-depth results of an RNA-Seq based comparison of our predictions. Our method does not yet produce comprehensive gene models but rather outputs base pair wise probabilities. Availability The source code of this work is available at https://github.com/weberlab-hhu/Helixer under the GNU General Public License v3.0. The trained models are available at https://doi.org/10.5281/zenodo.3974409 Supplementary information Supplementary data are available at Bioinformatics online.

Spliceator: multi-species splice site prediction using convolutional neural networks

“…Since SS and other regulatory motifs may be conserved across similar species [ 20 ], some work has been done to try to transfer models trained on model organisms to related organisms, for example between different vertebrate genomes [ 66 ]. Others have built cross-species models for specific clades, such as animals or plants using Helixer [ 67 ], but unfortunately the source code for this program is not yet stable (according to the authors). The aim of our work was to extend the idea of cross-species models to conceive universal SS prediction models (one for each SS), that are applicable to a wider range of organisms.…”

Transposable elements contribute to the establishment of the glycine shuttle in Brassicaceae species

“…Consistent structural gene annotations were generated for each species with Helixer (Stiehler et al 2020) using the hybrid convolutional and bidirectional long-short term memory model, , specifically the trained instance of . This was followed by post-processing the raw predictions into final primary gene models with Helixer Post (Bolger 2022, personal communication).…”

Helixer–de novoPrediction of Primary Eukaryotic Gene Models Combining Deep Learning and a Hidden Markov Model

“…Transition weights The base-wise probabilities output by the original version of Helixer [Stiehler et al, 2020] were scored against the references with the categorical cross-entropy loss function, and the penalty for each base pair varied at most by the setting of class weights (to compensate for class imbalance). This setup does not fully reflect the biological significance of different mistakes made by the network; and modifications to the Helixer architecture presented here address this.…”

“…For peceptibility, differences between the genic F1 of models are displayed as a Z-score (middle). Models marked with * used the same training species as inStiehler et al [2020]. Grey indicates data is not available.…”