TSS-Captur: A TSS-characterization pipeline for transcribed but unclassified RNA-transcripts

The aim of this project fork was to migrate TSS-Captur to the new Nextflow DSL2 syntax while integrating new features to improve functionality and user experience. In addition to the pipeline improvements, there was a strong focus on extending accessibility through the development of a web application. NextJS was chosen for its robust features and ease of integration with server-side operations.

To enable the characterization of transcripts in an explorative manner, we have developed TSS-Captur. TSS-Captur characterizes transcripts starting on clear TSS signals obtained from TSSpredator (Herbig, 2015) that cannot be associated to any known gene. To characterize the function of the transcript, TSS-Captur integrates tools using different approaches. The following figure describes the different steps and tools run by TSS-Captur.

It combines the advantages of comparative genomics and ab initio methods to account for many characteristics in the prediction. Furthermore, it uses the approach of transcriptional-signal based tools to combine TSS signals with a possible terminator. Moreover, it enables the user to explore the thermodynamic characteristics of possible identified ncRNA transcripts and visualize the secondary structure. Lastly, it facilitates the analysis of motifs within the promoter regions before each TSS signal to identify possible binding sites or other regulatory elements. All these results are combined within an interactive interface. This interface does not only provide an overview of the predicted results, but offers the possibility of getting the detailed information for each transcript and hence being able to reconstruct the predictions done by TSS-Captur.

How to run the pipeline

cd nf
./nextflow run tss_captur.nf --inputTable [PATH_TO_MASTER_TABLE] --inputGenomes [PATH_TO_GENOMES_DIR] --inputGFFs [PATH_TO_GFF_DIR] --outputDir [PATH_TO_OUTPUT] --blastdb [PATH_TO_NT_DB]  -with-docker mwittep/tsscaptur 

How to run the web application

Ensure Yarn Berry is installed, if not you can install it by following the instructions on Yarn documentation

cd nextjs
yarn install
yarn dev

This command will start the NextJS development server, making the web application accessible via http://localhost:3000 by default.

Acknowledgements

• Developed by Mathias Witte Paz, M.Sc. Bioinformatics, University of Tübingen
• Webified by Thomas Vogel, B.Sc. Bioinformatics, University of Tübingen