Evolution of virtual gene panels over time and implications for genomic data re-analysis

Abstract

Purpose: Re-analyzing genomic information from patients without a molecular diagnosis is known to improve diagnostic yields. There are different mechanisms responsible for this increase, but the discovery of new, and refinement of existing, gene-disease relationships are one of the most prominent drivers of new diagnoses. This study examines the incorporation of new knowledge into virtual diagnostic gene panels and how this affects the potential for re-analysis. Methods: We used PanelApp Australia to explore how the gene content of 112 rare-disease panels evolved between 2019 and 2022. By dividing these panels into groups that examined Specific and Broad rare-diseases clinical testing indications, we determined the granular changes in panel composition. Results: Characterizing how the panels present at the launch of PanelApp Australia changed, revealed that the diagnostic genes available for analysis increased in 82% of the Specific rare-disease panels and in 97% of the Broad rare-disease panels. Examining how the panels had evolved showed that different panels were changing at different rates and in different ways. The median number of diagnostic grade genes in the Specific rare-disease panel increased by 4 (0-63), whereas the median number of gene gains in the Broad rare-disease panels was 27 (0-432). Monthly snapshots demonstrated that these changes were highly variable among different panels. Conclusion: Knowledge about gene-disease associations is changing dynamically. Using fixed time periods may not be the best strategy to guide re-analysis frequency, as a result, some conditions may benefit from an approach based on the availability of new information rather than the passage of time.

Keywords

• Clinical genomics
• Diagnosis
• Disease genes
• Panelapp Australia
• Rare genetic disease
• Re-analysis of genomic data