G.A.M.E

G.A.M.E. accelerates elucidation of unknown chemical structures in natural product mixtures using high-resolution mass spectrometry data.

Key Features:

• GPU acceleration (NVIDIA CUDA): Implements GPU acceleration using NVIDIA's CUDA to increase computational throughput.
• Dynamic programming algorithm: Employs a dynamic programming approach to address the NP-complete problem of extending scaffold databases with sidechains that match mass data.
• Scaffold extension with sidechains: Extends scaffold databases by adding sidechains to generate candidate structures that can match observed masses.
• High-resolution mass support (five decimal digits): Handles high-resolution mass spectrometry data with precision up to five decimal digits.
• Parallel processing to mitigate exponential scaling: Uses parallel GPU processing to overcome the exponential increase in computation time associated with higher mass precision in DP algorithms.
• Matching to observed spectra: Matches extended scaffolds against observed high-resolution mass spectra to elucidate chemical structures present in mixtures.
• Validation on natural product datasets: Demonstrated on four datasets derived from natural products with verified compositions.

Scientific Applications:

• Metabolomics mixture analysis: Elucidates unknown compounds within complex natural product mixtures using mass spectrometry data.
• Structural elucidation in incomplete separation conditions: Identifies structures when chromatographic or instrumental separation does not fully resolve mixture components.
• Natural product discovery: Supports discovery and identification of new bioactive compounds from natural product extracts.
• High-precision MS data interpretation: Enables analysis and interpretation of mass spectrometry datasets requiring up to five decimal digits of mass precision.

Methodology:

Computations use NVIDIA CUDA GPU acceleration and a dynamic programming algorithm to extend scaffold databases with sidechains and match those extended scaffolds against observed high-resolution mass spectra, handling up to five decimal digits of precision and leveraging parallel processing to mitigate exponential runtime growth.