GenoSense is a robust, AI-driven genomics platform designed to streamline the flow of genetic data between hospitals, labs, and AI-powered analysis systems. It enables early disease detection, prediction of genetic vulnerabilities, and geographic visualization of common disease trends — all while maintaining strict data privacy through encryption.
This platform facilitates a complete pipeline for genomics-based healthcare prediction and disease mapping:
- Hospital Data Submission: Hospitals securely submit patient sample data to the system.
- Laboratory Processing: Labs receive sample requests and perform DNA sequencing.
- Variant Generation: Sequenced DNA is compared with the Human Genome Reference to extract genetic variants (VCF format).
- AI Vulnerability Analysis: The system uses a Hugging Face LLaMA model to analyze variants and predict possible health vulnerabilities.
- Human Insight Rendering: Predictions are interpreted and displayed in an easy-to-understand human-readable format.
- Geographic Disease Mapping: The system tracks common genetic conditions in nearby regions and displays them on an interactive map.
- End-to-End Encryption: All sensitive information (patient identity, DNA data, medical history) is encrypted during transit and storage.
- Submit patient samples to certified labs.
- Track lab request status.
- View AI-generated health risk predictions.
- Explore local disease hotspots based on genomic data.
- Accept or reject sequencing requests from hospitals.
- Upload DNA sequence data and generate VCF variant files.
- Monitor sample analytics and report predictions back to the hospital.
- Collaborate with AI models for high-accuracy genome analysis.
- Aggregates anonymized predictions from various hospitals/labs.
- Displays regions with high frequency of certain genetic vulnerabilities.
- Helps in identifying and preparing for possible public health concerns.
- All sample and patient data is encrypted end-to-end using secure cryptographic protocols.
- No raw DNA or personal identity is exposed during AI processing.
- Role-based access control ensures only authorized hospitals and labs interact with relevant data.
- Built using a fine-tuned LLaMA model from Hugging Face.
- Accepts VCF variant details and returns interpretable medical insights.
- Designed to aid genetic counselors, doctors, and patients in understanding potential health risks.
graph LR
A[Hospital] -->|Sample Sent| B[Lab]
B -->|DNA Sequencing| C[Generate VCF]
C --> D[AI Model: Predict Vulnerabilities]
D --> E[Hospital Receives Report]
D --> F[Map Common Vulnerabilities on Human Body and also marking common disease of a particular region on map]
- Frontend : Swelt
- Backend : Go
- Machine Learning : Flask , Python
Team Name: CodeNhiAta