D1.8 Tools Interoperability and Service Registry

The objective of EXCELERATE Deliverable 1.8, as stated in the original proposal, was to analyse the data format-usage landscape of tools registered the ELIXIR Tools and Data Services Registry (bio.tools) to provide a basis for targeted software developments to improve interoperability of registered tools. We originally foresaw those developments to be facilitated via a “Matchmaking Service mechanism” possibly including conversion of tools to use common formats, and development of format converter software where needed. Whilst such specific developments are laudable, it is now obvious that significant progress depends on the bioinformatics community at large, and requires vastly greater capacity than WP1 has at its disposal.

Instead, we took a much more targeted approach, focussing - as an exemplar - upon the systematic curation of tools for proteomics data analysis, and subsequent exploitation of the annotations for automated workflow composition. The work addresses the challenge of tool interoperability and is a major stride in delivering a practical “Matchmaking Service mechanism”. The work is described in two articles:

● Tsiamis, V., Ienasescu, H., Palmblad, M. Schwämmle, V. and Ison J. Community curation of software tools as illustrated for mass spectrometry based proteomics In preparation¨see https://tinyurl.com/proteomics-tools 

Palmblad, M., Lamprecht, A., Ison, J. and Schwämmle, V. (2018). Automated workflow composition in mass spectrometry based proteomics Accepted for publication in Bioinformatics 

The first article (Tsiamis et†al†.) describes a systematic approach towards the comprehensive coverage in bio.tools of the prevalent tools for proteomics data analysis, including expert curation of many tools to a high standard including consistent annotation of data formats and operation using the EDAM ontology. The second (Palmblad et al) explores automated workflow composition from the tool semantic annotation, and provides a toolkit to support researchers in identifying, comparing and benchmarking multiple workflows from individual bioinformatics tools.

For this deliverable report, we link to and summarise the publications in press. The project files and workflows are freely available:

● https://github.com/bio-tools/biotoolsCompose/tree/master/Automatic-Workflow-Co mposition