Plasmodesma

Plasmodesma performs dereplication and pharmacophoric deconvolution of complex plant extracts by analyzing nuclear magnetic resonance (NMR) spectroscopy datasets with machine learning to identify active compounds.

Key Features:

• Automated NMR Data Acquisition: Automates acquisition of a standard set of NMR experiments across medium-sized sample collections that vary by bioactivity to produce consistent spectral datasets.
• Machine Learning Integration: Applies machine learning algorithms to raw NMR data to extract spectral fingerprints indicative of active compounds.
• Pharmacophoric Deconvolution: Performs pharmacophoric deconvolution to isolate and characterize bioactive molecules within complex matrices by distinguishing their spectral signatures from inactive components.
• Optimal Condition Determination: Determines and validates optimal analysis conditions on a synthetic single-active-molecule sample and on a complex bioactive matrix exhibiting synergistic interactions.

Scientific Applications:

• Isolation and characterization of bioactive molecules: Identifies and characterizes active compounds from plant extracts using NMR-derived spectral fingerprints.
• Investigation of synergistic interactions: Deconvolves multi-component spectra to investigate synergistic interactions in complex natural matrices.
• Pharmacophore identification for drug discovery: Extracts pharmacophoric information from NMR data to support identification of potential pharmacophores in drug discovery workflows.

Methodology:

Uses a machine learning-driven analysis pipeline that extracts spectral fingerprints from raw NMR data and determines optimal analysis conditions validated on a synthetic single-active sample and a complex bioactive matrix.