VAREANT: a bioinformatics application for gene variant reduction and annotation.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• High-quality genomic variant data analysis is crucial for understanding the human genome, identifying biomarkers, and deciphering biological pathways.
• Despite increased genomic data availability, accessible bioinformatic tools for data preparation remain limited.
• Challenges in processing genomic data include large volumes, high computational costs, and difficulties in extracting relevant information.

Purpose of the Study:

• To present VAREANT, an accessible and configurable bioinformatic application designed to streamline the preparation of genomic variant data.
• To provide a tool that transforms raw variant data into an analysis-ready format, facilitating downstream research.

Main Methods:

• VAREANT consists of three modules: Pre-processing for data filtering, Variant Annotation for adding metadata, and AI/ML Data Preparation for creating machine-learning-ready datasets.
• The tool supports fine-grained filtering of complex variant datasets and integrates the latest public annotation databases.
• It is designed to create AI/ML-ready datasets with minimal user pre-processing.

Main Results:

• VAREANT has been successfully tested and validated on datasets of varying sizes.
• The application was implemented in two case studies involving patients with cardiovascular diseases.
• The tool effectively prepares genomic variant data for analysis, annotation, and machine learning applications.

Conclusions:

• VAREANT offers a comprehensive solution for preparing genomic variant data, addressing current accessibility and processing limitations.
• The application enhances the usability of genomic data for biomarker discovery, patient stratification, and pathway analysis.
• VAREANT is available as open-source code, promoting wider adoption and contribution within the research community.