Genomic databases are the bedrock of new moves towards more tailored and precise personalised medicine, and nowhere more so than in cancer. Tumour genome sequences can both inform diagnosis, prognosis and treatment for cancer, and serve as a crucial research resource to underpin the development of new and improved options for the same clinical elements.

Privacy is a contentious issue when it comes to genomics; on one hand, many individuals are very concerned about the potential for public exposure of their personal data; others are very much in favour of public sharing of such data to support research.

US firm Myriad Genetics has been making headlines for years thanks to the controversial gene patents it held for the BRCA 1/2 genes. Although aspects of these patents have now been revoked, opening up the US BRCA testing market to other companies, the company has retained dominance thanks to the massive database of BRCA mutations it has amassed over the years; this gives them a competitive edge in interpretation of results. However, a new public database is set to challenge this advantage as BRCA gene data accumulates in ClinVar, reportedly already containing 6,000 different BRCA variants (compared with some 16,000 in the Myriad version). Myriad has said that their own database is both of higher data quality and is not vulnerable to funding cuts, but proponents of the public database are advocating for patients, health professionals and health insurers to opt for BRCA-testing only via providers that are willing to share the DNA sequence data.

Moving outside the US to the international stage, the Global Alliance for Genomics and Health is developing plans for an international resource to link ClinVar and other large-scale, high quality clinical genetics databases.

Mean while, the International Cancer Genome Consortium (ICGC) has responded to renewed concerns about data privacy and the potential for re-identification of supposedly anonymous individuals with a review of their data sharing and confidentiality policies. Writing in Nature Biotechnology, they conclude that given the level of expertise in genetics and biostatistics needed for re-identification ‘do not create more than remote concerns’ for participants. Overall, they propose that ‘promises of absolute privacy and data protection are unrealistic’ in the age of genomic and other big data, but that rather researchers and policy-makers have obligations in terms of ‘vigilance, transparency and honesty’ to achieve a balance between protecting privacy and data accessibility.