Linking investigators

Many multifactorial diseases, such as various forms of cancer, diabetes, coronary heart disease and schizophrenia, still pose a considerable challenge for those researchers intent on understanding their causes. Some progress has been made in identifying the genes that contribute to such diseases and the availability of the sequence of the human genome holds further potential. However, scientists are still far from pinpointing the exact genetic causes of most major diseases that now plague humans in the developed world. Clinicians and epidemiologists have amassed a wealth of non‐genetic information, notably lifestyle and environmental factors, and with the rise of genomics, proteomics and imaging methods, banks holding tissue samples from patients have become an increasingly important and precious resource in medical research. However, the full value of all these efforts can only be realized by data sharing.A major problem is the fact that the data generated by researchers around the world are not easily available to those working in a related area. Although it has become considerably easier to identify relevant scientific publications, and to access gene or protein sequence data and analytical software, this unfortunately does not extend to primary data from studies and tissue banks. Indeed, scientists could gain valuable information if they had access to these primary data or to samples from their colleagues‘ studies. Current approaches, namely the pooling of diverse data in central databases and the formation of consortia between investigators, are fraught with the practical problems of data management, ownership of data and confidentiality. An alternative, as proposed in this Viewpoint, would be a centralized linking facility to issue anonymous ’linking numbers'. These would allow independent and autonomous research groups studying samples from the same subjects to be aware of each other, and to form voluntary and temporary collaborations for research on a specific problem. …