Bioinformatics for Human Proteomics: Current State and Future Status

Abstract

AbstractWe are entering a new era in the exploration of the human proteome.Advances in technologies are allowing researchers to map with a much better resolution and improved accuracy this very complex universe. Such an exploration also requires a compendium of adequate bioinformatics tools, data repositories and knowledge resources. We are going to describe what are the state of the art in term of data and knowledge resources for human proteins: from repositories of proteomics data such as PRIDE,PeptideAtlas or Peptidome, databases such as SRMAtlas and HPA and knowledgebase such as UniProtKB/Swiss-Prot.In September 2008, the UniProt/Swiss-Prot group achieved a major milestone: the first complete manual annotation of what is believed to be the full set of human proteins (derived from about 20'000 genes). This corpus of data is already quite rich in information pertinent to modern biomolecular medical research, but made us realize how large is the gap in our knowledge of human proteins in terms of functional information as well as protein characterization (PTMs, protein/protein interactions,subcellular locations, etc).This gap resides not only in the available experimental information, but also in the way this information has been stored, which is far from being sufficient to help researchers making sense of what all these human proteins do in our bodies! Therefore, in the framework of CALIPHO, a new interdisciplinary group created jointly by the University of Geneva and the Swiss Institute of Bioinformatics (SIB), we are developing neXtProt, a new human-centric protein knowledge resource.neXtProt is developed with the aim to help researchers answer pertinent questions relevant to human proteins. This requires answering 3 different challenges:1) Add to the corpus of data on human proteins that is already in Swiss-Prot, a lot of additional information. We will import in neXtProt data originating from a variety of high-throughput approaches such as microarray, antibodies, proteomics, siRNAs, interactomics, etc. All of these data sets must be carefully selected so as to only provide high-quality data as we want to avoid creating a noisy and dirty compendium.2) Organize the data in such a way that it is possible to make powerful queries in the most user-friendly environment. Here also, it is necessary to be able to capture the complexity and the heterogeneity of the data that will be available in neXtProt, yet make it easy for the user to forget this complexity!3) Build a software platform that will allows tools ranging from sequence analysis to text and data mining to be integrated in various research environments so as to answer specific needs of academic and industrial users.