Phenotype-driven gene target definition in clinical genome-wide sequencing data interpretation

Purpose: Genome-wide sequencing approaches are increasingly being used in place of disease gene panel sequencing approaches. Despite the well-recognized benefits of these approaches, they also carry with them an increased burden of analyzing overwhelmingly large gene targets and an increased possibility of detecting incidental findings. Methods: We propose a novel approach for design of individualized phenotype gene panels using the set of signs and symptoms observed and selecting relevant genes on the basis of known phenotype gene associations. Results: We used results of diagnostic exome sequencing in 405 cases submitted to our institution to show retrospectively that using the phenotype gene panel increases the sensitivity of masked exome analysis (increase from 25.4 to 29.7% in overall diagnostic yield). We also show that such a strategy enables the possibility of masked analysis of genome-wide sequencing data in patients with poorly defined and multifaceted clinical presentations. Ultimately, we show that this approach enables control over the incidental findings rate (0.25% in phenotype gene panels). Finally, we provide a Web tool for customized phenotype panel creation (available at http://www.kimg.eu/generator). Conclusion: In conclusion, we present a novel approach to a phenotype-driven diagnostic process of genome scale sequencing data that harnesses the sensitivity of these approaches while restricting the analysis to genes relevant to clinical presentation in patient.