Transposon annotation tool for the annotation of transposons, transposon characteristic proteins and structural elements of transposons. reasonaTE is part of TransposonUltimate.
The reasonaTE pipeline comes with two conda environments due to package incompatibilities. For some steps of the environment you will need the first, for others the second conda environment.
Installation using conda and mamba (recommended)
# Environment 1 - including all annotation tools
conda create -y --name transposon_annotation_tools_env python=2.7
conda activate transposon_annotation_tools_env
conda install -y mamba
#conda install -y -c bioconda repeatmodeler repeatmasker # Recommended not too install via conda
mamba install -y -c bioconda genometools-genometools # for some users: mamba install -y -c bioconda -c conda-forge genometools-genometools
mamba install -y -c derkevinriehl transposon_annotation_reasonate
mamba install -y -c derkevinriehl transposon_annotation_tools_proteinncbicdd1000
conda install -y -c derkevinriehl transposon_annotation_tools_transposonpsicli
mamba install -y -c derkevinriehl transposon_annotation_tools_mitetracker
mamba install -y -c derkevinriehl transposon_annotation_tools_sinescan=1.1.2
mamba install -y -c derkevinriehl transposon_annotation_tools_helitronscanner
mamba install -y -c derkevinriehl transposon_annotation_tools_mitefinderii
mamba install -y -c derkevinriehl transposon_annotation_tools_mustv2
mamba install -y -c derkevinriehl transposon_annotation_tools_sinefinder
mamba install -y -c anaconda biopython
conda deactivate
# Environment 2 - including CD-Hit and Transposon Classifier RFSB
conda create -y --name transposon_annotation_reasonaTE
conda activate transposon_annotation_reasonaTE
conda install -y mamba
mamba install -y -c anaconda biopython
mamba install -y -c bioconda cd-hit blast seqkit
mamba install -y -c derkevinriehl transposon_annotation_reasonate transposon_classifier_rfsb
conda deactivateInstallation using yml file (works for Linux64, other OS possible)
wget https://raw.githubusercontent.com/DerKevinRiehl/transposon_annotation_reasonaTE/main/environment_yml/transposon_annotation_tools_env.yml
wget https://raw.githubusercontent.com/DerKevinRiehl/transposon_annotation_reasonaTE/main/environment_yml/transposon_annotation_reasonaTE.yml
conda env create -f transposon_annotation_tools_env.yml
conda env create -f transposon_annotation_reasonaTE.ymlInstallation using plain conda (not recommended, can take long time)
# Environment 1 - including all annotation tools
conda create -y --name transposon_annotation_tools_env python=2.7
conda activate transposon_annotation_tools_env
#conda install -y -c bioconda repeatmodeler repeatmasker # Recommended not too install via conda
conda install -y -c bioconda genometools-genometools # for some users: conda install -y -c bioconda -c conda-forge genometools-genometools
conda install -y -c derkevinriehl transposon_annotation_reasonate
conda install -y -c derkevinriehl transposon_annotation_tools_proteinncbicdd1000
conda install -y -c derkevinriehl transposon_annotation_tools_transposonpsicli
conda install -y -c derkevinriehl transposon_annotation_tools_mitetracker
conda install -y -c derkevinriehl transposon_annotation_tools_sinescan=1.1.2
conda install -y -c derkevinriehl transposon_annotation_tools_helitronscanner
conda install -y -c derkevinriehl transposon_annotation_tools_mitefinderii
conda install -y -c derkevinriehl transposon_annotation_tools_mustv2
conda install -y -c derkevinriehl transposon_annotation_tools_sinefinder
conda install -y -c anaconda biopython
conda deactivate
# Environment 2 - including CD-Hit and Transposon Classifier RFSB
conda create -y --name transposon_annotation_reasonaTE
conda activate transposon_annotation_reasonaTE
conda install -y -c anaconda biopython
conda install -y -c bioconda cd-hit blast seqkit
conda install -y -c derkevinriehl transposon_annotation_reasonate transposon_classifier_rfsb
conda deactivateStep 1) Create a project
conda activate transposon_annotation_tools_env
mkdir workspace
wget https://raw.githubusercontent.com/DerKevinRiehl/transposon_annotation_reasonaTE/main/workspace/testProject/sequence.fasta # demo fasta you could use
reasonaTE -mode createProject -projectFolder workspace -projectName testProject -inputFasta sequence.fastaStep 2) Annotate genome with annotation tools To annotate the genome with different annotation tools, four possible ways exist. We recommend Option 2 as it allows for parallelization which is vital for reducing processing times for very large genomes.
Option 1: annotate with all tools automatically (this does not include ltrPred). This will annotate the genome with all tools (except for ltrPred) with standard parameters and tool after tool.
conda activate transposon_annotation_tools_env
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool allOption 2: annotate with one specific tool (good for parallelization or rerunning, recommended). It is mandatory to run the protein annotation tools transposonPSI and NCBICDD1000 for the next steps.
conda activate transposon_annotation_tools_env
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool helitronScanner
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool ltrHarvest
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool mitefind
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool mitetracker
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool must
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool repeatmodel
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool repMasker
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool sinefind
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool sinescan
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool tirvish
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool transposonPSI
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool NCBICDD1000Option 3: run annotation tools with specified parameters (for advanced users) If you want reasonaTE to call annotation tools with specific parameters, but do not want to take care of the locations of input and output files, you can do so as shown in the following example:
reasonaTE -mode annotate -projectFolder workspace -projectName testProject -tool tirvish xxxxx -mintsd 5Additional parameters need to follow after five x symbols "xxxxx". Please note, do only set parameters that are not related to locations of input and output files. If you want total control please have a look at Option 3.
Option 4: run annotation tools completely with user specified parameters (for expert users) For this purpose, we provide conda packages of all transposon_annotation_tools except for ltrPred. Please use the fasta file with renamed sequence names of the workspace project folder. (e.g. workspace/testProject/sequence.fasta) Please note, as some tools (HelitronScanner, MiteFinderII, MITE-Tracker, SINE-Finder, TIRvish) do not annotate on both strands, we recommend to run these on the reverse complementary as well (e.g. workspace/testProejct/sequence_rc.fasta). Once you annotated the genomes with your own specified parameter settings, please copy the result files into the workspace's project Folder as shown in the example project (e.g. results of HelitronScanner to copy into workspace/testProject/helitronScanner) and rename the files accordingly. Please note, it is mandatory to run the protein annotation tools transposonPSI and NCBICDD1000 for the next steps using the commands of option 2.
Running ltrPred: If you want to include ltrPred annotations into the pipeline as well, install and run ltrPred. Later on, copy the result files into the project folder (workspace/testProject/ltrPred) and rename the files accordingly. Please find our tutorial for manually running LTRpred even without docker using the conda package udocker. Based on our experience, ltrPred contributed valuable annotations including transposons and structure features. However, we were not able to create a conda package for easy and automated use, and it takes manual efforts to run it.
Check status of annotation tools: If you are running multiple annotation tools in parallel, or run the manually, copied and renamed the result files into the workspace folder, you can check the status of the annotation files by:
conda activate transposon_annotation_tools_env
reasonaTE -mode checkAnnotations -projectFolder workspace -projectName testProject
>Checking helitronScanner ... completed
>Checking ltrHarvest ... completed
>Checking ltrPred ... completed
>Checking mitefind ... completed
>Checking mitetracker ... completed
>Checking must ... completed
>Checking repeatmodel ... completed
>Checking repMasker ... completed
>Checking sinefind ... completed
>Checking sinescan ... completed
>Checking tirvish ... completed
>Checking transposonPSI ... completed
>Checking NCBICDD1000 ... completedAll files that are reported as "completed" will be considered by reasonaTE in the next steps.
Step 3) Parse annotations Each of the tools will produce different output file formats. reasonaTE therefore provides a parser module that will unify different output files to one standardized format (GFF3). The parser module will automatically detect annotations that are available as a result from step 2, and only the available files will be considered in the next steps by the pipeline.
conda activate transposon_annotation_tools_env
reasonaTE -mode parseAnnotations -projectFolder workspace -projectName testProjectIf you are unsure about the status of the parsing, you can run following command:
conda activate transposon_annotation_tools_env
reasonaTE -mode checkParsed -projectFolder workspace -projectName testProjectStep 4) Run the pipeline on the genome annotations
conda activate transposon_annotation_reasonaTE
reasonaTE -mode pipeline -projectFolder workspace -projectName testProjectStep 5) Calculate final statistics Once all results are calculated, summarizing statistics can be generated using:
conda activate transposon_annotation_reasonaTE
reasonaTE -mode statistics -projectFolder workspace -projectName testProjectThe results will be print to console and stored to the statistics files (see section "Documentation of output files" below). The results consist of three tables, presenting the number of transposons, the number of base pairs included by the transposon annotations and the number of base pairs annotated by the transposon mask annotation. The numbers are present by transposon class (horizontaly) and sequence (verticaly, using the renamed sequence names and the original sequence names) for the first two mentioned numbers, and just by sequences for the last mentioned number. All reported values are separated by tabulator. The three tables are separated by two empty lines.
SeqID SeqName #Num_transposons by classes
SeqID SeqName all 1 1/1 1/1/1 1/1/2 1/1/3 1/2 1/2/1 1/2/2 2 2/1 2/1/1 2/1/2 2/1/3 2/1/4 2/1/5 2/1/6 2/2 2/3
all all 28750 1290 1267 279 977 11 23 15 8 27460 27000 2276 4292 2952 1716 15674 90 77 383
seq1 chrI 3707 220 218 43 173 2 2 1 1 3487 3430 265 566 343 313 1935 8 10 47
seq2 chrII 5041 194 187 29 157 1 7 2 5 4847 4760 366 625 605 240 2908 16 7 80
...
SeqID SeqName #BP_transposons by classes
SeqID SeqName all 1 1/1 1/1/1 1/1/2 1/1/3 1/2 1/2/1 1/2/2 2 2/1 2/1/1 2/1/2 2/1/3 2/1/4 2/1/5 2/1/6 2/2 2/3
all all 30914788 5533640 5410363 951426 4438898 20039 123277 120973 2304 25381148 23998275 2331383 7151643 6435895 853227 7149583 76544 1097361 285512
seq1 chrI 4950902 1017833 1013925 204630 806578 2717 3908 3660 248 3933069 3356902 264572 1364087 688504 155671 876952 7116 520558 55609
seq2 chrII 4934076 583591 574266 102006 471723 537 9325 7763 1562 4350485 4192651 283217 1297024 901063 138632 1553229 19486 69580 88254
...
SeqID SeqName #BP_transposons
all all23412418
seq1 chrI 3875312
seq2 chrII 3492273
...| Usage Parameter Summary ModeNr | Mode | Parameter | Mandatory | Description |
|---|---|---|---|---|
| 1 | "createProject" | projectFolder | (mandatory) | Directory to create annotation projects in (=annotation workspace) |
| projectName | (mandatory) | Desired name of the annotation project | ||
| inputFasta | (mandatory) | Genome file (FASTA) that should be annotated for transposons | ||
| 2 | "annotate" | projectFolder | (mandatory) | Directory with annotation projects (=annotation workspace) |
| projectName | (mandatory) | Name of the annotation project | ||
| tool | (mandatory) | Annotation tool that should be used. Possible options: "helitronScanner", "ltrHarvest", "mitefind", "mitetracker", "must", "repeatmodel", "repMasker", "sinefind", "sinescan", "tirvish", "transposonPSI", "NCBICDD1000", "all" | ||
| 3 | "checkAnnotations" | projectFolder | (mandatory) | Directory with annotation projects (=annotation workspace) |
| projectName | (mandatory) | Name of the annotation project | ||
| 4 | "parseAnnotations" | projectFolder | (mandatory) | Directory with annotation projects (=annotation workspace) |
| projectName | (mandatory) | Name of the annotation project | ||
| 5 | "checkParsed" | projectFolder | (mandatory) | Directory with annotation projects (=annotation workspace) |
| projectName | (mandatory) | Name of the annotation project | ||
| 6 | "pipeline" | projectFolder | (mandatory) | Directory with annotation projects (=annotation workspace) |
| projectName | (mandatory) | Name of the annotation project | ||
| 7 | "statistics" | projectFolder | (mandatory) | Directory with annotation projects (=annotation workspace) |
| projectName | (mandatory) | Name of the annotation project | ||
| 8 | "sequenceRenamer" | seqNames | (mandatory) | sequence_heads.txt file location with original and new sequence names |
| inputGFF | (mandatory) | Input GFF file | ||
| outputGFF | (mandatory) | Target location of GFF file with renamed (=original) sequences |
Introduction The outputs of the pipeline consist of mainly two parts:
Project folder structure Inside a project's folder (e.g. testProject) there are multiple output folders, that are presented in the following. The collapsed folders and marked files (by the + symbol in green) represent the relevant output files:
+├── finalResults
+│ ├── FinalAnnotations_ProteinFeatures.gff3
+│ ├── FinalAnnotations_StructuralFeatures.gff3
+│ ├── FinalAnnotations_TransposonMask.gff3
+│ ├── FinalAnnotations_TransposonSequences.fasta
+│ ├── FinalAnnotations_Transposons.gff3
+│ ├── PipelineAnnotations_ProteinFeatures.gff3
+│ ├── PipelineAnnotations_TransposonMask.gff3
+│ ├── PipelineAnnotations_TransposonSequences.fasta
+│ ├── PipelineAnnotations_Transposons.gff3
+│ ├── ToolAnnotations_ProteinFeatures.gff3
+│ ├── ToolAnnotations_StructuralFeatures.gff3
+│ ├── ToolAnnotations_TransposonMask.gff3
+│ ├── ToolAnnotations_TransposonSequences.fasta
+│ └── ToolAnnotations_Transposons.gff3
├── helitronScanner
├── helitronScanner_rc
├── ltrHarvest
├── ltrPred
├── mitefind
├── mitefind_rc
├── mitetracker
├── mitetracker_rc
├── must
├── NCBICDD1000
+├── parsedAnnotations
+│ ├── helitronScanner.fasta
+│ ├── helitronScanner.gff3
+│ ├── ltrHarvest.fasta
+│ ├── ltrHarvest.gff3
+│ ├── ltrPred.fasta
+│ ├── ltrPred.gff3
+│ ├── mitefind.fasta
+│ ├── mitefind.gff3
+│ ├── mitetracker.fasta
+│ ├── mitetracker.gff3
+│ ├── must.fasta
+│ ├── must.gff3
+│ ├── NCBICDD1000.gff3
+│ ├── proteinfeatures.gff3
+│ ├── proteinfeatures_masked2.gff3
+│ ├── proteinfeatures_masked3.gff3
+│ ├── proteinfeatures_masked.gff3
+│ ├── repeatmodel.fasta
+│ ├── repeatmodel.gff3
+│ ├── repeatmodel_repeats.gff3
+│ ├── repMasker.fasta
+│ ├── repMasker.gff3
+│ ├── repMasker_repeats.gff3
+│ ├── sinefind.fasta
+│ ├── sinefind.gff3
+│ ├── sinescan.fasta
+│ ├── sinescan.gff3
+│ ├── tirvish.fasta
+│ ├── tirvish.gff3
+│ └── transposonPSI.gff3
├── repeatmodel
├── repMasker
+├── sequence.fasta
+├── sequence_heads.txt
+├── sequence_rc.fasta
├── sinefind
├── sinefind_rc
├── sinescan
+├── Statistics_FinalAnnotations.txt
+├── Statistics_ToolAnnotations.txt
├── tirvish
├── tirvish_rc
├── transposonCandA
├── transposonCandB
├── transposonCandC
├── transposonCandD
├── transposonCandE
├── transposonCandF
└── transposonPSIFirst of all, the fasta file used for the creation of the project was copied to sequence.fasta. The sequences in the fasta file were renamed, a matching can be found in sequence_heads.txt. Also, the reverse complement sequence was copied to sequence_rc.fasta for all softwares that annotate a single strand only. If you would like to use the original sequence names, you can do so using Mode 8 of reasonaTE (see table before).
Moreover, Statistics_FinalAnnotations.txt and Statistics_ToolAnnotations.txt contain the statistics produced by the statistics mode for the two outputs of reasonaTE.
The folder parsedAnnotations includes the parsed transposon annotations, structural feature annotations and transposon characteristic protein annotations by the different software tools in GFF3 format, as well as extracted sequences for each annotation in a FASTA file.
The folder finalResults includes all results - including the tool and pipeline annotations. The ToolAnnotations_ files contain the tool annotations, the PipelineAnnotations_ files contain the additional copies found and the FinalAnnotations_ include both of the prior merged into one file. There are files of the annotated transposons, transposon characteristic proteins, structural features, the mask of transposon regions and the extracted and classified sequences as FASTA file. As transposon annotations are not intersection free and can include nested or overlapping transposon annotations, the basepairs annotated in the mask represent all base pairs that are annotated by one or more transposons of the transposon annotations.
Please cite our paper if you find TransposonUltimate useful:
Kevin Riehl, Cristian Riccio, Eric A Miska, Martin Hemberg, TransposonUltimate: software for transposon classification, annotation and detection, Nucleic Acids Research, 2022; gkac136, https://doi.org/10.1093/nar/gkac136
@article{riehl2022transposonultimate,
title={TransposonUltimate: software for transposon classification, annotation and detection},
author={Riehl, Kevin and Riccio, Cristian and Miska, Eric and Hemberg, Martin},
journal={Nucleic Acids Research},
year={2022}
}We would like to thank Sarah Buddle, Simone Procaccia, Fu Xiang Quah and Alexandra Dallaire for their assistance with testing and debugging the software.