Obtaining novel biomedical knowledge from proteomics research

Bioinformatics (the use of computer science to analyze biological data) is now an essential element in the field of biomedicine, and with ever-improving instrumentation and continuous growth and complexity of the generated data, the need constantly increases. The field of proteomics (the analysis of protein expression) is no exception, and one of the current major bioinformatics-related bottlenecks is the interpretation of the results in a larger biological context. Research projects often come to a halt after acquiring a list of proteins, without knowing how to interpret the project's findings in light of existing knowledge. And while the amount of available information is growing, the focus in the research community has mainly been on the gathering of the information and not on making it available to other researchers in an interactive manner. This has resulted in the paradox that even though the sharing of proteomics data and knowledge has been standardized and simplified, the gathered information has not to the same extent been made easily available for biomedical applications, and therefore remains largely unexploited. It is therefore clear that the way in which public proteomics data previously have being made available is not compatible with the goals of most biomedical research projects. This project worked on improving the interactive integration of project-specific proteomics results with existing biomedical knowledge from online repositories. This has made it easier for biomedical researchers to obtain novel biomedical knowledge, and inspire novel experimental research directions building on their own proteomics data. The project depended on two main sources of information: in-house generated proteomics data, and publicly available biomedical knowledge. The in-house generated data, focused on experimental studies related to multiple sclerosis, has been made available via the freely available database CSF-PR (proteomics.uib.no). Moving beyond single protein biomarkers to protein networks and biological pathways was also an essential task, thus helping to improve the understanding of how the interesting proteins interact in a bigger biological context.

Prosjektet har resultert i utviklingen av CSF-PR (proteomics.uib.no/csf-pr), en brukervennlig nettbasert database som gjør det enkelt å få oversikt over potensielle proteinmarkører for nevrologiske sykdommer som multiple sklerose (MS), Alzheimer, Parkinson og ALS. Databasen er basert på proteomikkdata hentet fra fagfellevurderte vitenskapelige publikasjoner, inkludert egenproduserte forskningsdata. Den samlede informasjonen i CSF-PR har deretter blitt brukt til å utvikle målrettede analyser for kvantifikasjon av de mest lovende proteinmarkørene i nytt pasientmaterialet. Håpet er at slike analyser over tid vil gjøre det enklere å sammenligne data fra ulike forskningsgrupper. Noe som vil kunne bidra til en bedre forståelse av forskjellene i proteinuttrykning mellom de ulike nevrologiske sykdommene, og i det lange løp forhåpentligvis komme pasientene til gode i form av bedre behandling.

Bioinformatics is now an essential element in the field of biomedicine, and with ever improving instrumentation and continuous growth and complexity of the generated data, the need constantly increases. The field of mass spectrometry based proteomics research is no exception, and one of the current major bioinformatics related bottlenecks is the interpretation of the results in a larger biomedical context. Research projects often come to a halt after acquiring a list of identified and/or quantified proteins, without knowing how to interpret the project's findings in light of existing knowledge. And while the amount of available data is growing, the focus in the research community has mainly been on the gathering of the data and not on making it available to the researchers in an interactive manner. This has resulted in the paradox that even though the sharing of proteomics data has been standardized and simplified, the data has in some ways become less available for biomedical applications, and therefore remains largely unexploited. Indeed, researchers are often simply overwhelmed by the sheer amount of information. It is therefore clear that the way in which public proteomics data are currently made available is not compatible with the goals of most biomedical research projects. The main objective of the this project is to improve the interactive integration of project-specific proteomics results with existing biomedical knowledge from online repositories. This will ensure that biomedical researchers will obtain novel biomedical knowledge, and will moreover inspire novel experimental research directions that build on their proteomics data. The project will depend on two main sources of information: in-house generated proteomics data, and publicly available biomedical knowledge. The in-house generated data, focused on experimental studies related to multiple sclerosis and diabetes, will serve to test and to validate the developed bioinformatics approaches.