CAGI 2013: Assessing the State of Genome Interpretation

CAGICritical Assessment of Genome Interpretation (CAGI) meeting in Berlin, where we shared our insights on how best to interpret human genomes, and, in turn, learned from expert colleagues about their own methods and data sources.

Each year, CAGI challenges the scientific community to interpret genome datasets from people with and without particular diseases, and taps independent assessors to systematically score how well the entrants predict who’s sick and why.

Participants then gather to discuss the results, helping us gauge how far science has come — and how far we have to go! — on the road to making genomes reliably informative in healthcare. Top-scoring entrants, along with assessors and organizers, explain which methods worked best, prompting thoughtful open discussion of breakthroughs and remaining hurdles.

This year, Ingenuity entered two CAGI challenges. One presented exomes from a family of four, and asked entrants to guess who among them shows far too little so-called good cholesterol (HDL-C). The other challenge presented exomes from another family, with more complex cholesterol and triglyceride problems, and asked entrants to guess who among them shows which symptom(s) — and, in turn, to pick out the sequence variant(s) most likely to blame.

We used Ingenuity Variant Analysis to tackle both these challenges, and were delighted by how the platform performed. In the HDL-C challenge, we scored best among the 39 entries. And in the complex cholesterol/triglyceride challenge, we were one of just three entries, again among 39, to find the putative causal variant.

Our interpretations strongly leveraged two key facets of the Variant Analysis platform: our built-in functional knowledge (of which genes interactively drive relevant diseases and other body processes) and our handy genome comparison features (which, for each CAGI challenge, let us easily wrap in comprehensive genome data from other challenges as extra controls).

In Berlin, we greatly enjoyed connecting with other scientific minds, whose diverse insights are essential to refining genome interpretation methods. CAGI is a vital initiative, and we commend the organizers for gathering key stakeholders to evaluate interpretation methods, and lay groundwork for future standards in the field.

As such, we’re very excited to work further with CAGI, and with fellow participants, to help make genome data a cornerstone of clinical decision-making for future generations