Development of Bioinformatics Tools and Infrastructure for Quantitative Proteomics

The most important result of the project has been the establishment of a user-friendly and freely available platform for the analysis of proteomics data. A platform that can be used both to carry out the analysis and to visualize and validate the results. The platform supports both protein identification and protein quantification and is freely available online at http://peptide-shaker.googlecode.com. Another central goal has been to set up structured and organized storage of the large amounts of data generated in a common proteomics lab. This includes both local storage for daily use and long term storage of raw data for possible future re-analysis. For the latter the collaboration with the Computational Biology Unit at the University of Bergen has be en utilized to develop a system for automatic backup using StoreBioinfo (https://storebioinfo.norstore.no). While the local structured storage is supported by the development of a LIMS-system (http://colims.googlecode.com) developed together with partners at the University of Ghent, Belgium. A focus on user-friendliness and open-source has been essential in the development of all the software. Both elements that makes it easier for other research groups to start using the developed tools, and if wanted contribute to the further development. The hope is that this will increase the re-use of existing code and knowledge from previous projects, both nationally and internationally. Another key element in the project has been the dissemination of the resu lts and the education of other researchers and students in the use of freely available software for the analysis of proteomics data. In this context extensive teaching material has been developed and used and at numerous international courses in bioinform atics for proteomics. All of the material is also available online: http://compomics.com/bioinformatics-for-proteomics. In total, parts of the project has been presented at 13 national and international courses and conferences. Supervision of two PhD stu dents, two Master students and three project students also highlights the project's focus on teaching. The project has so far resulted in 12 scientific publications in international journals, plus several additional manuscripts currently "under review" or planed submitted in the near future. International collaboration has also been a central element in the project with collaborators in both Belgium, Great Britain and Germany. Together with national collaborators at the University of Bergen and Haukela nd University Hospital, this has increased the quality of the developed systems. A four month's stay as a guest researcher in Belgium at the University of Ghent has also helped strengthen the international bond between the two research groups.

The identification and quantification of proteins using proteomics has rapidly become one of the most important areas of protein research. Especially within the context of clinical quantitative proteomics, the ability to simultaneously detect and quantify hundreds or even thousands of proteins is developing into a very promising field with a widespread set of possibilities. Quantitative mass spectrometry based proteomics is however a new field in rapid development. Large amounts of data are generated from such experiments, and there are very few standardized solutions for data interpretation and data organization. In addition there is a high number of different analysis methods used and most of them require tailor-made software solutions. Yet another comp licating factor is that each instrument vendor uses its own data format. Together this results in a demand for novel software solutions that will make it possible to analyze proteomics data in an efficient and correct way, a step that currently is a bottl eneck for any proteomics laboratory worldwide. Making a robust but flexible infrastructure for storing, analyzing and sharing the proteomics data will therefore be the main focus in the project proposed here. The conduction of these objectives will be tig htly connected to the proteomics platform at the University of Bergen (PROBE), and apart from directly yielding publishable tools and algorithms, this will also ensure that the state of the art proteomics methods developed and in use at PROBE will be matc hed by equally well-adapted bioinformatics solutions. This will in particular benefit all projects that use PROBE and the NorProteomics consortium as well as other proteomics laboratories in Norway. The project will also have a high international impact w hich is also secured through our collaboration with international pioneers in the field.