SimMS

SimMS accelerates cosine similarity calculations for tandem mass spectrometry through GPU implementation, enabling repository-scale spectral library matching for compound identification.

Key Features:

• GPU Acceleration: SimMS achieves approximately 1000-fold speedup compared to the MatchMS reference implementation.
• Accuracy Preservation: SimMS provides this speedup without any loss of accuracy.
• Repository-Scale Matching: SimMS enables spectral library matching at repository scale without the need for large compute clusters.
• Broad Applicability: SimMS extends to any spectral comparison-based methods including molecular networking approaches and analogue search.

Scientific Applications:

• Compound Identification: SimMS supports large-scale comparison of fragmentation patterns of mass spectra against databases containing known spectra for compound identification in untargeted metabolomics.
• Molecular Networking: SimMS accelerates spectral comparison-based molecular networking approaches.
• Analogue Search: SimMS enables spectral comparison-based analogue search methods.

Methodology:

SimMS implements GPU-accelerated cosine similarity for tandem mass spectrometry.