Best use of experimental data in cancer informatics.

The welcome attitude of the 'omics community, journals and funders of research towards data sharing, coupled with successful implementations of data standards, has resulted in resource dissemination and a better understanding of many diseases, including cancer. Sharing experiment data is beneficial in terms of knowledge generation, allowing reproduction and validation of results. An adherence to a reporting guideline enables full-value extraction from costly data; this is an inexpensive method to increased quality without incurring disproportionate costs. For therapy data in particular, easy access to the range of new approaches and the ability to perform valid comparisons between these approaches would be especially useful. We discuss initiatives that support resource sharing and summarize three reporting guidelines for experiment data that have been adopted successfully. Finally, we introduce a new guideline that encompasses the diverse data types in therapeutic experiments, which is intended to be of use to the cancer therapeutics community.