Box 2 Biobanks

The efficient mining of large collections of clinical and epidemiological data requires the availability of electronic and standardized records coupled to organized collections of samples in biological banks (biobanks). The concept of a biobank covers efforts with different goals and organization, from efforts to obtain samples from the general population, to collections dedicated to specific diseases, in particular cancer types. Biobanks also vary greatly in the type of sample-associated information they contain. In some cases this comprises very detailed clinical and epidemiological records, and in others only basic descriptions of population characteristics. At a very general level, three main types of biobanks can be distinguished [60].

Population biobanks gather germline DNA from healthy donors representing a particular regional population. Their major goal is to obtain biomarkers of susceptibility and population characteristics.

Disease-oriented biobanks focus on the identification of disease biomarkers for patient selection. They store collections of pathological and healthy samples commonly associated with clinical data or trials. Well-known examples are tumor biobanks.

Epidemiology-oriented biobanks focus on exposure biomarkers. Samples are recruited from healthy exposed individuals or from case-control studies.

Current efforts in biobank development include the European Biobanking and Biomolecular Resources Infrastructure (BBMRI), which intends to coordinate biobanks from 19 European countries, including the organization of compatible infrastructures and annotations [61]. The European Life-sciences Infrastructure for Biological Information project (ELIXIR [61]), another project of BBMRI, represents an effort to link biological and biomedical databases and computational resources [61]. In the same way, the NCI Biomedical Informatics Grid project (CaBIG) supports the integration of medical oncology and cancer research genome projects [62]. The Public Population Project in Genomics consortium (P3G) gathers together more than 20 international institutions to promote effective collaborations between biobanks involved in population studies [63].

Examples of specific biobank developments are the Estonian Gene Bank Project [64], the private initiative of deCODE project in Iceland [65], the Spanish National Tumor Bank network [66], the DNA scanning project in children, in the Children's Hospital of Philadelphia (CHOP) [67], the Personalized Medicine Research Project DNA Biobank [68] in the United States, and the BioBank Japan Project [69].