Predicting Collision-Induced-Dissociation\nTandem Mass\nSpectra (CID-MS/MS) Using Ab Initio Molecular Dynamics

Compound identification is at the center of metabolomics,\nusually\nby comparing experimental mass spectra against library spectra. However,\nmost compounds are not commercially available to generate library\nspectra. Hence, for such compounds, MS/MS spectra need to be predicted.\nMachine learning and heuristic models have largely failed except for\nlipids. Here, quantum chemistry software can be used to predict mass\nspectra. However, quantum chemistry predictions for collision induced\ndissociation (CID) mass spectra in LC-MS/MS are rare. We present the\nCIDMD (Collision-Induced Dissociation via Molecular Dynamics) framework\nto model CID-based MS/MS spectra. It uses first-principles molecular\ndynamics (MD) to simulate the physical process of molecular collisions\nin CID tandem mass spectrometry. First, molecular ions are constructed\nat specific protonation sites. Using density functional theory, these\nprotonated ions are targeted by argon collider gas atoms at user-specified\nvelocities. Subsequent bond breakages are simulated over time for\nat least 1,000 fs. Each simulation is repeated multiple times from\nvarious collisional directions. Fragmentations are accumulated over\nthose repeated collisions to generate CIDMD in silico mass spectra.\nTwelve small metabolites (<205 Da) were selected to test the accuracy\nof this framework in comparison to experimental MS/MS spectra. When\ntesting different protomers, collider velocities, number of simulations,\nsimulation time and impact factor b cutoffs, we yielded 261 predicted\nmass spectra. These in silico spectra resulted in entropy similarity\nscores of an average 624 ± 189 for all 261 spectra compared to\ntheir corresponding experimental spectra, which improved to 828 ±\n77 when using optimal parameters of the most probable protomers for\n12 molecules. With increasing molecular mass, higher velocities achieved\nbetter results. Similarly, different protomers showed large differences\nin fragmentation; hence, with increasing numbers of protomers and\ntautomers, the average CIDMD prediction accuracy decreased. Mechanistic\ndetails showed that specific fragment ions can be produced from different\nprotomers via multiple fragmentation pathways. We propose that CIDMD\nis a suitable tool to predict mass spectra of small metabolites like\nproduced by the gut microbiome.