May 19, 2014
Isaac Kohane, David Margulies and colleagues have reported on the CLARITY Challenge in Genome Biology, which sought to develop standards for the analysis, interpretation and reporting of clinical sequencing for the diagnosis of genetic disorders. The study challenge included information on 12 individuals from 3 families with different heritable disorders, with clinical information provided from medical records, exome sequencing using the sOliD platform and whole-genome sequencing by complete genomics. Details of the families included in the challenge are described in the video above.
Thirty international groups participated in the challenge and were assessed on the basis of the methods they used for analysis and interpretation, whether these methods were efficient, scalable and replicable, and the clinical usefulness of their case reports. Only two groups identified the consensus candidate variants in all of the cases, although there was significant overlap in the candidates reported across the teams and some consensuses emerging on the methods used for alignment, variant calling and pathogenicity prediction. For example, most of the groups used gatK and saMtools, either alone or in combination, for variant calling and used both siFt and PolyPhen for pathogenicity prediction. The authors highlight the need for broader adoption of standard data formats, consideration of coverage along with an estimation of false negative rates for candidate genes and further development of publicly available genomic databases.