Rapid and Automated Ab Initio Metabolite Collisional Cross Section Prediction from SMILES Input

We developed a novel approach for predicting collisional cross sections (CCS) using an ab initio workflow. This workflow incorporates various techniques, including molecular mechanics, deep learning potentials, conformational clustering, and quantum mechanics (QM). By automating these steps in a high-performance computing (HPC) environment, users can input the SMILES structure of small organic molecules and obtain a Boltzmann averaged CCS value as the output. CCS is a metric measured by ion mobility spectrometry (IMS) and can greatly enhance the analysis of untargeted metabolomics experiments. In our study, we found that replacing the QM geometry refinement step with a single-point energy calculation resulted in only a slight decrease in accuracy, enabling us to expedite the CCS calculation. Despite involving stochastic steps such as conformation generation and clustering, the final CCS value remained highly reproducible. Additionally, our research demonstrates that the gas phase ensembles modeled in this workflow can serve as intermediate files for predicting other properties, such as aqueous phase nuclear magnetic resonance chemical shift prediction. For those interested, the software used in this study is available on GitHub at the following link: github.com/DasSusanta/snakemake_ccs.