Annotation Of Eukaryotic Genomes Research Articles

Welcome to the big leaves: Best practices for improving genome annotation in non-model plant genomes.

Robust standards to evaluate quality and completeness are lacking in eukaryotic structural genome annotation, as genome annotation software is developed using model organisms and typically lacks benchmarking to comprehensively evaluate the quality and accuracy of the final predictions. The annotation of plant genomes is particularly challenging due to their large sizes, abundant transposable elements, and variable ploidies. This study investigates the impact of genome quality, complexity, sequence read input, and method on protein-coding gene predictions. The impact of repeat masking, long-read and short-read inputs, and de novo and genome-guided protein evidence was examined in the context of the popular BRAKER and MAKER workflows for five plant genomes. The annotations were benchmarked for structural traits and sequence similarity. Benchmarks that reflect gene structures, reciprocal similarity search alignments, and mono-exonic/multi-exonic gene counts provide a more complete view of annotation accuracy. Transcripts derived from RNA-read alignments alone are not sufficient for genome annotation. Gene prediction workflows that combine evidence-based and ab initio approaches are recommended, and a combination of short and long reads can improve genome annotation. Adding protein evidence from de novo assemblies, genome-guided transcriptome assemblies, or full-length proteins from OrthoDB generates more putative false positives as implemented in the current workflows. Post-processing with functional and structural filters is highly recommended. While the annotation of non-model plant genomes remains complex, this study provides recommendations for inputs and methodological approaches. We discuss a set of best practices to generate an optimal plant genome annotation and present a more robust set of metrics to evaluate the resulting predictions.

AMAW: automated gene annotation for non-model eukaryotic genomes

Background: The annotation of genomes is a crucial step regarding the analysis of new genomic data and resulting insights, and this especially for emerging organisms which allow researchers to access unexplored lineages, so as to expand our knowledge of poorly represented taxonomic groups. Complete pipelines for eukaryotic genome annotation have been proposed for more than a decade, but the issue is still challenging. One of the most widely used tools in the field is MAKER2, an annotation pipeline using experimental evidence (mRNA-seq and proteins) and combining different gene prediction tools. MAKER2 enables individual laboratories and small-scale projects to annotate non-model organisms for which pre-existing gene models are not available. The optimal use of MAKER2 requires gathering evidence data (by searching and assembling transcripts, and/or collecting homologous proteins from related organisms), elaborating the best annotation strategy (training of gene models) and efficiently orchestrating the different steps of the software in a grid computing environment, which is tedious, time-consuming and requires a great deal of bioinformatic skills. Methods: To address these issues, we present AMAW (Automated MAKER2 Annotation Wrapper), a wrapper pipeline for MAKER2 that automates the above-mentioned tasks. Importantly, AMAW also exists as a Singularity container recipe easy to deploy on a grid computer, thereby overcoming the tricky installation of MAKER2. Use case: The performance of AMAW is illustrated through the annotation of a selection of 32 protist genomes, for which we compared its annotations with those produced with gene models directly available in AUGUSTUS. Conclusions: Importantly, AMAW also exists as a Singularity container recipe easy to deploy on a grid computer, thereby overcoming the tricky installation of MAKER2

New insights into genome annotation in Podospora anserina through re-exploiting multiple RNA-seq data

BackgroundPublicly available RNA-seq datasets are often underused although being helpful to improve functional annotation of eukaryotic genomes. This is especially true for filamentous fungi genomes which structure differs from most well annotated yeast genomes. Podospora anserina is a filamentous fungal model, which genome has been sequenced and annotated in 2008. Still, the current annotation lacks information about cis-regulatory elements, including promoters, transcription starting sites and terminators, which are instrumental to integrate epigenomic features into global gene regulation strategies.ResultsHere we took advantage of 37 RNA-seq experiments that were obtained in contrasted developmental and physiological conditions, to complete the functional annotation of P. anserina genome. Out of the 10,800 previously annotated genes, 5’UTR and 3’UTR were defined for 7554, among which, 3328 showed differential transcriptional signal starts and/or transcriptional end sites. In addition, alternative splicing events were detected for 2350 genes, mostly due alternative 3’splice sites and 1732 novel transcriptionally active regions (nTARs) in unannotated regions were identified.ConclusionsOur study provides a comprehensive genome-wide functional annotation of P. anserina genome, including chromatin features, cis-acting elements such as UTRs, alternative splicing events and transcription of non-coding regions. These new findings will likely improve our understanding of gene regulation strategies in compact genomes, such as those of filamentous fungi. Characterization of alternative transcripts and nTARs paves the way to the discovery of putative new genes, alternative peptides or regulatory non-coding RNAs.

A chromosome-level genome assembly reveals genomic characteristics of the American mink (Neogale vison)

Availability of a contiguous chromosome-level genome assembly is the foundational step to develop genome-based studies in American mink (Neogale vison). The main objective of this study was to provide a high quality chromosome-level genome assembly for American mink. An initial draft of the genome assembly was generated using 2,884,047 PacBio long reads. Integration of Hi-C data into the initial draft led to an assembly with 183 scaffolds and scaffold N50 of 220 Mb. This gap-free genome assembly of American mink (ASM_NN_V1) had a length of 2.68 Gb in which about 98.6% of the whole genome was covered by 15 chromosomes. In total, 25,377 genes were predicted across the American mink genome using the NCBI Eukaryotic Genome Annotation Pipeline. In addition, gene orthology, demographic history, synteny blocks, and phylogenetic relationships were studied in connection with the genomes of other related Carnivora. Furthermore, population-based statistics of 100 sequenced mink were presented using the newly assembled genome. Remarkable improvements were observed in genome contiguity, the number of scaffolds, and annotation compared to the first draft of mink genome assembly (NNQGG.v01). This high-quality genome assembly will support the development of efficient breeding strategies as well as conservation programs for American mink.

The dark proteome: translation from noncanonical open reading frames.

Omics-based technologies have revolutionized our understanding of the coding potential of the genome. In particular, these studies revealed widespread unannotated open reading frames (ORFs) throughout genomes and that these regions have the potential to encode novel functional (micro-)proteins and/or hold regulatory roles. However, despite their genomic prevalence, relatively few of these noncanonical ORFs have been functionally characterized, likely in part due to their under-recognition by the broader scientific community. The few that have been investigated in detail have demonstrated their essentiality in critical and divergent biological processes. As such, here we aim to discuss recent advances in understanding the diversity of noncanonical ORFs and their roles, as well as detail biologically important examples within the context of the mammalian genome.

The genome sequence of the common green lacewing, Chrysoperla carnea (Stephens, 1836).

We present a genome assembly from an individual female Chrysoperla carnea (a common green lacewing; Arthropoda; Insecta; Neuroptera; Chrysopidae). The genome sequence is 560 megabases in span. The majority of the assembly (95.70%) is scaffolded into six chromosomal pseudomolecules, with the X sex chromosome assembled. Gene annotation of this assembly by the NCBI Eukaryotic Genome Annotation Pipeline has identified 12,985 protein coding genes.

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

BackgroundAb initio prediction of splice sites is an essential step in eukaryotic genome annotation. Recent predictors have exploited Deep Learning algorithms and reliable gene structures from model organisms. However, Deep Learning methods for non-model organisms are lacking.ResultsWe developed Spliceator to predict splice sites in a wide range of species, including model and non-model organisms. Spliceator uses a convolutional neural network and is trained on carefully validated data from over 100 organisms. We show that Spliceator achieves consistently high accuracy (89–92%) compared to existing methods on independent benchmarks from human, fish, fly, worm, plant and protist organisms.ConclusionsSpliceator is a new Deep Learning method trained on high-quality data, which can be used to predict splice sites in diverse organisms, ranging from human to protists, with consistently high accuracy.

The genome sequence of the common frog, Rana temporaria Linnaeus 1758.

We present a genome assembly from an individual female Rana temporaria (the common frog; Chordata; Amphibia; Anura; Ranidae). The genome sequence is 4.11 gigabases in span. The majority of the assembly is scaffolded into 13 chromosomal pseudomolecules. Gene annotation of this assembly by the NCBI Eukaryotic Genome Annotation Pipeline has identified 23,707 protein coding genes.

The genome sequence of the common toad, Bufo bufo (Linnaeus, 1758).

We present a genome assembly from an individual male Bufo bufo (the common toad; Chordata; Amphibia; Anura; Bufonidae). The genome sequence is 5.04 gigabases in span. The majority of the assembly (99.1%) is scaffolded into 11 chromosomal pseudomolecules. Gene annotation of this assembly by the NCBI Eukaryotic Genome Annotation Pipeline has identified 21,517 protein coding genes.

Chromosome-Level Reference Genome Assembly for the American Pika (Ochotona princeps).

The American pika (Ochotona princeps) is an alpine lagomorph found throughout western North America. Primarily inhabiting talus slopes at higher elevations (>2000 m), American pikas are well adapted to cold, montane environments. Warming climates on both historical and contemporary scales have contributed to population declines in American pikas, positioning them as a focal mammalian species for investigating the ecological effects of climate change. To support and expand ongoing research efforts, here, we present a highly contiguous and annotated reference genome assembly for the American pika (OchPri4.0). This assembly was produced using Dovetail de novo proximity ligation methods and annotated through the NCBI Eukaryotic Genome Annotation pipeline. The resulting assembly was chromosome- scale, with a total length of 2.23 Gb across 9350 scaffolds and a scaffold N50 of 75.8 Mb. The vast majority (>97%) of the total assembly length was found within 36 large scaffolds; 33 of these scaffolds correlated to whole autosomes, while the X chromosome was covered by 3 large scaffolds. Additionally, we identified 17 enriched gene ontology terms among American pika-specific genes putatively related to adaptation to high-elevation environments. This high-quality genome assembly will serve as a springboard for exploring the evolutionary underpinnings of behavioral, ecological, and taxonomic diversification in pikas as well as broader-scale eco-evolutionary questions pertaining to cold-adapted species in general.

Growing and Sustaining a Nationwide Course‐based Undergraduate Research Experience (CURE): Genomics Education Partnership Enhances Research Opportunities at Diverse Institutions

The Genomics Education Partnership (GEP) is a community of practice focused on integrating genomics research experiences into undergraduate curriculum and supporting faculty in this endeavor. Founded in 2006, GEP has grown to over 100 participating two-year and four-year institutions and introduced thousands of students to eukaryotic gene structure, comparative genomics, and genome evolution. GEP members contribute their expertise and enthusiasm to the community through four committees: science and IT, curriculum development, assessment, and professional development and mentoring. Digital communication tools facilitate collaboration in this geographically distributed community while regional nodes enable the development of local initiatives. Our education research is focused on student experiences, attitudes, and learning gains in the CURE setting. An important dimension of CUREs that distinguishes them from the traditional laboratory classroom is iteration: students have opportunities to make mistakes and re-do (Auchincloss et al, 2014). The scaling of undergraduate research from an apprentice model to the classroom, however, poses a major challenge for faculty, creating the need to provide feedback to a large number of inexperienced student researchers. We find that custom bioinformatics tools provide rapid feedback to GEP students and allow for iterative revisions of student projects. Student surveys and focus groups reveal that bioinformatics CUREs foster “formative frustration,” whereby students can safely fail in their original analysis, adjust, recover, and succeed; students value this experience. During the COVID pandemic, GEP has provided an authentic research experience that can be delivered remotely. Bioinformatics resources are available online, as are GEP virtual tutors. Thus, GEP is a viable option for undergraduate research projects even for online learning. We are expanding our repertoire of scientific projects. GEP has partnered with Galaxy to develop G-OnRamp, an open-source platform for constructing UCSC Assembly Hubs and JBrowse/Apollo genome browsers to facilitate collaborative annotation of eukaryotic genomes. These tools allow us to include investigations of the evolution of venom in parasitoid wasps and an investigation of the evolution of genes in Drosophila biochemical pathways. We seek to grow our network of faculty and to increase the diversity of our community by making a variety of training opportunities available, including online mentoring via the QUBES platform and regional node workshops. We have created a supportive community of faculty and welcome new colleagues (no experience necessary!) eager for the challenge of bringing genomics research to undergraduate students at all levels. Faculty interested in joining the GEP can contact us at http://gep.wustl.edu/contact_us.

InsectOR-Webserver for sensitive identification of insect olfactory receptor genes from non-model genomes.

Insect Olfactory Receptors (ORs) are diverse family of membrane protein receptors responsible for most of the insect olfactory perception and communication, and hence they are of utmost importance for developing repellents or pesticides. Accurate gene prediction of insect ORs from newly sequenced genomes is an important but challenging task. We have developed a dedicated webserver, 'insectOR', to predict and validate insect OR genes using multiple gene prediction algorithms, accompanied by relevant validations. It is possible to employ this server nearly automatically and perform rapid prediction of the OR gene loci from thousands of OR-protein-to-genome alignments, resolve gene boundaries for tandem OR genes and refine them further to provide more complete OR gene models. InsectOR outperformed the popular genome annotation pipelines (MAKER and NCBI eukaryotic genome annotation) in terms of overall sensitivity at base, exon and locus level, when tested on two distantly related insect genomes. It displayed more than 95% nucleotide level precision in both tests. Finally, given the same input data and parameters, InsectOR missed less than 2% gene loci, in contrast to 55% loci missed by MAKER for Drosophila melanogaster. The webserver is freely available on the web at http://caps.ncbs.res.in/insectOR/ and the basic package can be downloaded from https://github.com/sdk15/insectOR for local use. This tool will allow biologists to perform quick preliminary identification of insect olfactory receptor genes from newly sequenced genomes and also assist in their further detailed annotation. Its usage can also be extended to other divergent gene families.

BRAKER2: automatic eukaryotic genome annotation with GeneMark-EP+ and AUGUSTUS supported by a protein database.

The task of eukaryotic genome annotation remains challenging. Only a few genomes could serve as standards of annotation achieved through a tremendous investment of human curation efforts. Still, the correctness of all alternative isoforms, even in the best-annotated genomes, could be a good subject for further investigation. The new BRAKER2 pipeline generates and integrates external protein support into the iterative process of training and gene prediction by GeneMark-EP+ and AUGUSTUS. BRAKER2 continues the line started by BRAKER1 where self-training GeneMark-ET and AUGUSTUS made gene predictions supported by transcriptomic data. Among the challenges addressed by the new pipeline was a generation of reliable hints to protein-coding exon boundaries from likely homologous but evolutionarily distant proteins. In comparison with other pipelines for eukaryotic genome annotation, BRAKER2 is fully automatic. It is favorably compared under equal conditions with other pipelines, e.g. MAKER2, in terms of accuracy and performance. Development of BRAKER2 should facilitate solving the task of harmonization of annotation of protein-coding genes in genomes of different eukaryotic species. However, we fully understand that several more innovations are needed in transcriptomic and proteomic technologies as well as in algorithmic development to reach the goal of highly accurate annotation of eukaryotic genomes.

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

MotivationCurrent 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.ResultsWe 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 and implementationThe 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.3974409Supplementary information Supplementary data are available at Bioinformatics online.

OpenProt 2021: deeper functional annotation of the coding potential of eukaryotic genomes.

OpenProt (www.openprot.org) is the first proteogenomic resource supporting a polycistronic annotation model for eukaryotic genomes. It provides a deeper annotation of open reading frames (ORFs) while mining experimental data for supporting evidence using cutting-edge algorithms. This update presents the major improvements since the initial release of OpenProt. All species support recent NCBI RefSeq and Ensembl annotations, with changes in annotations being reported in OpenProt. Using the 131 ribosome profiling datasets re-analysed by OpenProt to date, non-AUG initiation starts are reported alongside a confidence score of the initiating codon. From the 177 mass spectrometry datasets re-analysed by OpenProt to date, the unicity of the detected peptides is controlled at each implementation. Furthermore, to guide the users, detectability statistics and protein relationships (isoforms) are now reported for each protein. Finally, to foster access to deeper ORF annotation independently of one’s bioinformatics skills or computational resources, OpenProt now offers a data analysis platform. Users can submit their dataset for analysis and receive the results from the analysis by OpenProt. All data on OpenProt are freely available and downloadable for each species, the release-based format ensuring a continuous access to the data. Thus, OpenProt enables a more comprehensive annotation of eukaryotic genomes and fosters functional proteomic discoveries.

Growing and Sustaining a Nationwide CURE: Genomics Education Partnership Enhances Research Opportunities for Students and Faculty at Diverse Institutions

The Genomics Education Partnership (GEP) is a community of practice focused on the integration of genomics research experiences into undergraduate curriculum and supporting faculty in this endeavor. Founded in 2006, GEP has grown to 100+ participating institutions and introduced thousands of students to eukaryotic gene structure, comparative genomics, and genome evolution. Recently, GEP has transitioned to a collaborative leadership structure. GEP members contribute their expertise and enthusiasm to the community through four committees: science and IT, curriculum development, assessment, and professional development and mentoring. Digital communication tools facilitate collaboration in this geographically distributed community and allow faculty with similar interests to work together. We seek to grow our network of faculty and to increase the diversity of our community by making a variety of training opportunities available, including online mentoring via the QUBES platform and regional node workshops. We are also expanding our repertoire of scientific projects. GEP has partnered with Galaxy to develop G‐OnRamp, an open‐source platform for constructing UCSC Assembly Hubs and JBrowse/Apollo genome browsers to facilitate collaborative annotation of eukaryotic genomes. These tools allow us to expand GEP projects, for example, to include investigations of the evolution of venom in parasitoid wasps and an investigation of the evolution of insulin pathway genes across 27 Drosophila genomes. Our education research is focused on student experiences, attitudes, and learning gains in the Course‐based Undergraduate Research Experience (CURE) setting. An important dimension of CUREs that distinguishes them from the traditional laboratory classroom is iteration: students have opportunities to make mistakes and redo (Auchincloss et al, 2014). The scaling of undergraduate research from an apprentice model to the classroom, however, poses a major challenge for faculty, creating the need to provide feedback to a large number of inexperienced student researchers. We find that custom bioinformatics tools provide rapid feedback to GEP students and allow for iterative revisions of student projects. Student surveys and focus groups reveal that bioinformatics CUREs foster “formative frustration,” whereby students can safely fail in their original analysis, adjust, recover, and succeed; students value this experience. We have created a supportive community of faculty and welcome new colleagues (no experience necessary!) eager for the challenge of bringing genomics research to all undergraduate students (freshmen through seniors). Faculty interested in joining the GEP can contact us at http://gep.wustl.edu/contact_us.Support or Funding InformationSupported by NSF IUSE‐1915544 and NIH IPERT‐1R25GM130517‐01 to LKR.

G-OnRamp: a Galaxy-based platform for collaborative annotation of eukaryotic genomes.

SummaryG-OnRamp provides a user-friendly, web-based platform for collaborative, end-to-end annotation of eukaryotic genomes using UCSC Assembly Hubs and JBrowse/Apollo genome browsers with evidence tracks derived from sequence alignments, ab initio gene predictors, RNA-Seq data and repeat finders. G-OnRamp can be used to visualize large genomics datasets and to perform collaborative genome annotation projects in both research and educational settings.Availability and implementationThe virtual machine images and tutorials are available on the G-OnRamp web site (http://g-onramp.org/deployments). The source code is available under an Academic Free License version 3.0 through the goeckslab GitHub repository (https://github.com/goeckslab).Supplementary information Supplementary data are available at Bioinformatics online.

VAPiD: a lightweight cross-platform viral annotation pipeline and identification tool to facilitate virus genome submissions to NCBI GenBank

BackgroundWith sequencing technologies becoming cheaper and easier to use, more groups are able to obtain whole genome sequences of viruses of public health and scientific importance. Submission of genomic data to NCBI GenBank is a requirement prior to publication and plays a critical role in making scientific data publicly available. GenBank currently has automatic prokaryotic and eukaryotic genome annotation pipelines but has no viral annotation pipeline beyond influenza virus. Annotation and submission of viral genome sequence is a non-trivial task, especially for groups that do not routinely interact with GenBank for data submissions.ResultsWe present Viral Annotation Pipeline and iDentification (VAPiD), a portable and lightweight command-line tool for annotation and GenBank deposition of viral genomes. VAPiD supports annotation of nearly all unsegmented viral genomes. The pipeline has been validated on human immunodeficiency virus, human parainfluenza virus 1–4, human metapneumovirus, human coronaviruses (229E/OC43/NL63/HKU1/SARS/MERS), human enteroviruses/rhinoviruses, measles virus, mumps virus, Hepatitis A-E Virus, Chikungunya virus, dengue virus, and West Nile virus, as well the human polyomaviruses BK/JC/MCV, human adenoviruses, and human papillomaviruses. The program can handle individual or batch submissions of different viruses to GenBank and correctly annotates multiple viruses, including those that contain ribosomal slippage or RNA editing without prior knowledge of the virus to be annotated. VAPiD is programmed in Python and is compatible with Windows, Linux, and Mac OS systems.ConclusionsWe have created a portable, lightweight, user-friendly, internet-enabled, open-source, command-line genome annotation and submission package to facilitate virus genome submissions to NCBI GenBank. Instructions for downloading and installing VAPiD can be found at https://github.com/rcs333/VAPiD.

Peptimapper: proteogenomics workflow for the expert annotation of eukaryotic genomes

BackgroundAccurate structural annotation of genomes is still a challenge, despite the progress made over the past decade. The prediction of gene structure remains difficult, especially for eukaryotic species, and is often erroneous and incomplete. We used a proteogenomics strategy, taking advantage of the combination of proteomics datasets and bioinformatics tools, to identify novel protein coding-genes and splice isoforms, assign correct start sites, and validate predicted exons and genes.ResultsOur proteogenomics workflow, Peptimapper, was applied to the genome annotation of Ectocarpus sp., a key reference genome for both the brown algal lineage and stramenopiles. We generated proteomics data from various life cycle stages of Ectocarpus sp. strains and sub-cellular fractions using a shotgun approach. First, we directly generated peptide sequence tags (PSTs) from the proteomics data. Second, we mapped PSTs onto the translated genomic sequence. Closely located hits (i.e., PSTs locations on the genome) were then clustered to detect potential coding regions based on parameters optimized for the organism. Third, we evaluated each cluster and compared it to gene predictions from existing conventional genome annotation approaches. Finally, we integrated cluster locations into GFF files to use a genome viewer. We identified two potential novel genes, a ribosomal protein L22 and an aryl sulfotransferase and corrected the gene structure of a dihydrolipoamide acetyltransferase. We experimentally validated the results by RT-PCR and using transcriptomics data.ConclusionsPeptimapper is a complementary tool for the expert annotation of genomes. It is suitable for any organism and is distributed through a Docker image available on two public bioinformatics docker repositories: Docker Hub and BioShaDock. This workflow is also accessible through the Galaxy framework and for use by non-computer scientists at https://galaxy.protim.eu.Data are available via ProteomeXchange under identifier PXD010618.

Long-Read Annotation: Automated Eukaryotic Genome Annotation Based on Long-Read cDNA Sequencing.

Single-molecule full-length complementary DNA (cDNA) sequencing can aid genome annotation by revealing transcript structure and alternative splice forms, yet current annotation pipelines do not incorporate such information. Here we present long-read annotation (LoReAn) software, an automated annotation pipeline utilizing short- and long-read cDNA sequencing, protein evidence, and ab initio prediction to generate accurate genome annotations. Based on annotations of two fungal genomes (Verticillium dahliae and Plicaturopsis crispa) and two plant genomes (Arabidopsis [Arabidopsis thaliana] and Oryza sativa), we show that LoReAn outperforms popular annotation pipelines by integrating single-molecule cDNA-sequencing data generated from either the Pacific Biosciences or MinION sequencing platforms, correctly predicting gene structure, and capturing genes missed by other annotation pipelines.