Fig. 1: Representative clinical workflow to uncover disease-causing genetic variants in undiagnosed patients.

Upon acceptance to the Undiagnosed Diseases Network (UDN), (a) an affected patient has an in-person clinical evaluation where extensive phenotyping and additional tests are performed as needed. (b) Before or during the clinical evaluation, samples of relevant affected and unaffected individuals in a family are sent for genomic sequencing. (c,d) Sequencing data provided by the sequencing center are analyzed in conjunction with other information in a back-and-forth process between bioinformaticians, clinicians, and genetic counselors to highlight variants that are likely to explain the patient’s disease. (e) Matches to the strong candidate explanatory variants identified in (c) are searched for in databases containing human genetic variant and corresponding symptom information (e.g., Matchmaker Exchange) or in databases containing animal genetic variants and corresponding phenotype information (e.g., MARRVEL). Strong candidate variants are also introduced into model organisms or cell lines where possible to assess in vivo phenotypic impact. (f) Once a candidate variant has been confirmed as disease causal, a molecular diagnosis is provided that can subsequently be used to tailor clinical management and molecular therapeutics. (g–j) Recurring steps in computational workflows to process genomic sequencing data to call, filter, and prioritize genetic variants that explain the affected individual’s disease symptoms.