Summary Points * An increasing number of public/private initiatives are exploring novel ways of conducting scientific research, including the use of social media and online collection of self-reported data. * Research relying on collection of self-reported data by self-selected participants has known methodological limitations, including selection bias, information bias, and confounding. * Such limitations may mean that results and conclusions of research using data obtained through online communities need to be interpreted with caution, as further replication is often required. * The findings of research, including their potential actionability, should be communicated to participants in a way that is understandable, accurate, complete, and not misleading. * The potential for sharing participants' data with third parties as well as the commercial uses of research findings should be disclosed to participants prior to consent. Using self-reported phenotypic data provided by their customers, 23andMe reported that they replicated over 180 genetic associations from the catalogue of genomewide association studies (GWAS) of the National Human Genome Research Institute's Office of Population Genomics 3, identified genetic associations for miscellaneous traits long suspected of having a genetic basis 4, and identified two novel loci and a substantial genetic component for Parkinson disease 5.