This platform provides a centralized hub for variant effect prediction, serving as a one-stop resource to assess the potential impact of genetic mutations on protein function. It compiles comprehensive data on human genomic variants and integrates multiple tools for efficient and insightful analysis.
The primary variant database was provided by the University of Mozambique. It includes all genomic positions with possible variants, enriched with detailed annotations:
Due to the massive size of the full genome dataset, the platform has been optimized for speed. By default, only variants related to three key proteins — collagen, keratin, and elastin — are preloaded. This allows users to perform rapid queries and obtain results almost instantly for common searches.
Users can also search for variants outside of the three default proteins. These extended searches access the full genome database and may take longer to complete. To enhance user experience, such queries are handled asynchronously. The system performs the analysis in the background and sends an email to the user once the results are ready.
For each variant, the platform provides prediction scores from established algorithms to estimate the functional impact. Current tools include:
The application backend uses structured tables to manage users, predictions, and variant results:
USERS: stores user credentials and session dataPrediction: links methods and scores to user queriesdbNSFP_Results: main variant data table with annotations and prediction outputsAll searches are recorded and linked to the user's account, ensuring traceability and personalized access to prior results.