CoSIMS: An Optimized Trajectory-Based Collision Simulator for Ion Mobility Spectrometry.
Christopher A MyersRebecca J D'EspositoDaniele FabrisSrivathsan V RanganathanAlan A ChenPublished in: The journal of physical chemistry. B (2019)
A new, multithreaded, trajectory method based software platform, CoSIMS, is revealed and compared to reference MOBCAL collision cross sections (CCS). CoSIMS employs various molecular mechanics algorithms to lessen the computational resources required to simulate thousands of buffer gas-ion collisions, including the neglect of London dispersion interactions at long distances and the removal of trajectories that insignificantly contribute to the total CCS via an ellipsoidal projection approximation. The showcased program is used to calculate the collision cross sections of carbon fullerenes, proteins, and DNA strands of various lengths, sizes, and molecular weights, and these are compared against the CCSs calculated by MOBCAL. Through this analysis, it is shown that the application of the aforementioned algorithms enables both faster and more reasonable CCS calculations than MOBCAL for highly elongated molecules such as nucleic acids; for all other molecules, CoSIMS is able to reproduce the CCSs generated by MOBCAL's trajectory method within a few percent. Overall, CoSIMS is able to calculate nearly identical CCSs as MOBCAL in nearly 2 orders of magnitude less CPU time due to the various numerical methods implemented into the software, even when run on a single CPU core.
Keyphrases
- machine learning
- single molecule
- deep learning
- data analysis
- high resolution
- depressive symptoms
- circulating tumor
- molecular dynamics
- quality improvement
- molecular dynamics simulations
- high throughput
- cell free
- single cell
- computed tomography
- magnetic resonance
- image quality
- virtual reality
- carbon dioxide
- contrast enhanced