TREND

TREND analyzes multivariate biophysical data to track equilibrium and nonequilibrium population shifts among two-dimensional data frames.

Key Features:

• Pattern Discovery: Applies Principal Component Analysis (PCA) directly to densely populated spectra, microscopy images, and dynamic movies without prior peak or feature selection to identify equilibrium states and temporal changes.
• Independent Component Analysis (ICA): Leverages Independent Component Analysis alongside PCA to uncover independent underlying signal components in multivariate datasets.
• Concurrent Process Identification: Detects concurrent processes within datasets, exemplified by extracting dual binding phases from NMR spectra during titrations and resolving breathing and cardiac components in MRI movies.
• Data Reconstruction: Reconstructs series of measurements using selected principal components, with reconstruction fidelity dependent on the number of PCs used.
• Format Compatibility: Reads spectra from spectroscopies in JCAMP-DX and NMR formats and ingests multiple movie formats for imaging and dynamic data analysis.

Scientific Applications:

• NMR Spectroscopy: Extracts binding phases and analyzes titration dynamics from NMR-detected titrations.
• MRI Imaging: Resolves physiological cycles such as breathing and cardiac activity from cardiac MRI movies.
• General Biophysical Measurements: Identifies equilibrium and temporal changes across spectra, microscopy images, and dynamic movie datasets.

Methodology:

Applies PCA and ICA to densely populated spectral, imaging, and movie datasets without prior peak/feature selection; performs component-based reconstruction using selected principal components; reads input data from JCAMP-DX, NMR formats, and multiple movie formats.