IMSDB

IMSDB integrates, validates, and analyzes metabolic and volatile organic compound (VOC) datasets—including ion mobility spectrometry with multi-capillary columns (MCC/IMS) chromatograms capable of detecting VOCs at low concentrations in moist exhaled air—to support biomarker identification for pulmonary diseases from human breath.

Key Features:

• Centralized Data Repository: Combines metabolic maps with heterogeneous datasets such as patient records and MCC/IMS chromatograms into a cohesive structure for biomarker identification.
• Hybrid Database Design: Implements a hybrid entity-attribute-value (EAV) model extended with EAV-CR concepts incorporating classes and relationships to organize data and improve retrieval efficiency.
• Automated Data Integration and Validation: Provides automated processes for integrating new data into the repository with validation mechanisms for high-throughput MCC/IMS datasets.
• Versioning and Roll-back Strategy: Supports version control to track changes over time and revert to previous data states to maintain consistency.
• Semi-Automatic Data Mining and Machine Learning: Incorporates semi-automated data mining and machine learning techniques to explore large datasets and identify patterns and potential biomarkers.
• Support for Biomarker Identification and Validation: Tailors data structures and analysis workflows to support identification and validation of biomarkers using MCC/IMS technology for pulmonary disease diagnostics.

Scientific Applications:

• Respiratory clinical research: Characterization and analysis of VOCs in human breath using MCC/IMS data to investigate pulmonary disease-related signatures.
• Biomarker discovery and validation: Identification and validation of candidate biomarkers for pulmonary disease diagnostics and potential personalized treatment strategies.
• Translational integrative studies: Integration of metabolic maps with patient records and MCC/IMS chromatograms for translational biomarker studies.

Methodology:

Uses a hybrid EAV/EAV-CR database architecture, automated data integration and validation processes, version control with rollback, semi-automatic data mining and machine learning techniques, and integration of metabolic maps with heterogeneous datasets including patient records and MCC/IMS chromatograms.