Development of computational approaches for the analysis of proteomic data.
Protein analysis by mass spectrometry (MS) went through big technological improvements in the past decade, in a manner almost similar to developments in genomic research. However, the development of computational approaches targeting analysis of MS data lagged behind. This is particularly relevant regarding meta-analysis, i.e., the combined analysis of MS data collected independently by different research groups. Such type of analysis is relevant because, while data collected for a specific project can fulfill only narrow objectives (for example, a project studying membrane fractions from a cell of the immune system), meta-analysis can explore wider questions (using the same example above, the analysis of many datasets where different fractionation methods were used can help to evaluate and optimize the efficiency of such methods). The worst bottleneck on this type of analysis is how to handle the normalization of the data that was collected by different groups using unique instruments and approaches. This raises the need for computational tools using mathematical approaches that will allow for that. Therefore, the main aim of this project is to employ computational biology in the proteomics field, being able to normalize and compare samples collected independently and which can be applied to several topics.