VUStruct: A compute pipeline for high throughput and personalized structural biology.

Area of Science:

• Genomics
• Computational Biology
• Structural Biology

Background:

• Interpreting genetic variants of unknown significance (VUSs) is crucial for diagnosing rare genetic disorders.
• Current diagnostic methods rely on gene-phenotype databases and DNA scoring, which may lack depth.
• The increasing availability of 3D protein structures, sequencing data, and AI necessitates advanced analytical tools.

Purpose of the Study:

• To introduce VUStruct, a web-accessible variant analysis pipeline, designed to supplement existing methods for VUS interpretation.
• To leverage 3D protein structure analysis for a deeper understanding of the molecular impact of genetic variations.
• To provide a mature, extensible, High-Performance Computing (HPC) software pipeline for both clinicians and researchers.

Main Methods:

• VUStruct maps missense variants onto protein 3D structures.
• It performs energy-based assessments of protein folding and stability.
• Includes pathogenicity prediction, ML-based analysis of binding surface disruptions, and identification of digenic disease genes.

Main Results:

• VUStruct has been successfully applied to over 175 Undiagnosed Disease Network (UDN) patient cases.
• The pipeline's analyses have informed clinical decision-making and guided additional patient testing.
• VUStruct has also informed research directions in computational genomics and wet lab studies for collaborators.

Conclusions:

• VUStruct offers a powerful, structure-based approach to VUS interpretation, enhancing rare disease diagnosis.
• The pipeline integrates multiple analytical methods to provide comprehensive insights into variant impact.
• VUStruct serves as a valuable tool for both clinical rare disease genome interpretation and academic research.