chronic vs clinic

chronic vs clinic performs conversion of heterogeneous genomic sequencing data into image representations and applies machine learning, including Support Vector Machine classification, to infer viral infection stage and detect transmission clusters from next-generation sequencing of viral quasispecies such as HCV HVR1.

Key Features:

• Novel Preprocessing Approach: Converts complex genomic datasets into normalized image data to enable application of image-classification methods.
• Machine Learning Integration: Reframes genomic analysis as an image-classification problem and applies machine learning models, including Support Vector Machines, for classification and comparison of diverse datasets.
• Molecular Epidemiology Focus: Targets inference of infection stages and detection of transmission clusters and outbreaks using next-generation sequencing data.

Scientific Applications:

• Inference of Viral Infection Stages: Applied to Hepatitis C Virus (HCV) hypervariable region 1 (HVR1) samples from recently and chronically infected individuals, achieving over 95% accuracy in infection staging.
• Detection of Transmission Clusters and Outbreaks: Validated on data from 33 epidemiologically curated outbreaks with accuracy exceeding 97% for clustering and outbreak detection.

Methodology:

Genomic data are transformed into image representations; machine learning image-classification models, including Support Vector Machine-based approaches, are used for infection-stage classification; clustering is performed on epidemiologically curated outbreak data.