A clinically driven variant prioritization framework outperforms purely computational approaches for the diagnostic...

Area of Science:

• Genomics
• Clinical Genetics
• Bioinformatics

Background:

• Next-generation sequencing (NGS) requires efficient variant identification for clinical application.
• The Melbourne Genomics Health Alliance (MGHA) developed a variant prioritization framework incorporating gene and variant indices.
• Previous methods relied heavily on computational variant properties.

Purpose of the Study:

• To evaluate the MGHA variant prioritization framework's effectiveness in ranking causative variants from singleton whole exome sequencing (WES) data.
• To compare the MGHA framework's performance against other gene and variant prioritization tools.
• To assess the impact of clinician-generated gene lists on variant prioritization accuracy.

Main Methods:

• Utilized data from 80 patients undergoing singleton WES.
• Applied the MGHA variant prioritization framework, including gene prioritization index and variant prioritization index (VPI).
• Compared the framework's performance against other computational prioritization tools and analyzed the contribution of clinician-generated gene lists.

Main Results:

• Causative variants were identified in 59 out of 80 patients.
• The MGHA framework ranked causative variants highly, with an average rank of 2.24 and 90% within the top five.
• Clinician-generated gene lists improved causative variant ranking by an average of 8.2 positions compared to variant properties alone, outperforming computational tools (P=0.001).

Conclusions:

• Clinically led variant prioritization significantly enhances the efficiency of singleton WES data analysis.
• The MGHA framework demonstrates superior performance compared to purely computational approaches.
• These findings support the development of clinically integrated models for genomic service delivery and funding.