Sharing of scientific data has many benefits. It boosts research, speeds
up translation and helps ensure that good practice is maintained. The teams that
generate the data are rewarded too: studies show that sharing data increases
citation rates.

The Wellcome Trust Sanger
Institute in the UK, a key player in the Human Genome Project, has often led
the way in this area, and in the latest issue of Genome Medicine, Tim Hubbard and Stephanie Dyke from the Wellcome
Trust Sanger Institute explain  how they
developed and implemented the Institute’s policy.

The Human Genome Project was groundbreaking in many ways, one of which
was the decision, known as the “Bermuda Agreement” or “Bermuda Principles”, to
release data as soon as possible and well ahead of publication. Over the years,
 early data sharing became established as
an import aspect of genomic research.

The success
of the Human Genome Project led to the launch of many new initiatives and
studies based on genomic science and  different types of data. In addition, more and
more human data were used and this had implications for confidentiality.  A consensus was reached around 2006-2007 that
the way forward was to share data in a managed way that limited access to the
data for approved purposes. In recognition of this, the Wellcome Trust Sanger
Institute began consulting on its own data-sharing policy.

Tim Hubbard
and colleagues approached the task by breaking it into three steps: guidance,
facilitation and oversight. The aim of the guidance part was to understand the
types of data being generated, the quality required to be useful to colleagues
and the timelines involved. During the facilitation stage, the barriers to
sharing – technical (such as the time required to get the data in the correct
format) and concerning credit for the work – were addressed.

The final
step required was oversight; a way of monitoring data sharing was needed and a
governance body, the data-sharing working group, was established.

As the
field of genomic medicine moves on, probably with the integration of electronic
health records with genetic data, policies for data sharing will evolve and
need to be updated. The three steps followed by the Wellcome Trust Sanger
Institute are likely to provide a framework for other institutes implementing
their own policy.