Deep learning for retention time prediction in reversed-phase liquid chromatography.

Retention time prediction in high-performance liquid chromatography (HPLC) is the subject of many studies since it can improve the identification of unknown molecules in untargeted profiling using HPLC coupled with high-resolution mass spectrometry. Lots of approaches were developed for retention time prediction in liquid chromatography for a different number of molecules considering various molecular properties and machine learning algorithms. The recently built large retention time data set of standard compounds from the Metabolite and Chemical Entity Database (METLIN) allows researchers to create a model that can be used for retention time prediction of small molecules with wide varieties of structures and physicochemical properties. The ability to predict retention times using the largest data set was studied for different architectures of deep learning models that were trained on molecular fingerprints, and SMILES (string representation of a molecule) represented as one-hot matrices. The best result was achieved with a one-dimensional convolutional neural network (1D CNN) that uses SMILES as an input. The proposed model reached the mean absolute error and the median absolute error equal to 34.7 and 18.7 s, respectively, which outperformed the results previously obtained for this data set. The pre-trained 1D CNN on the METLIN SMRT data set was transferred on five other data sets to evaluate the generalization ability.

Omics Untargeted Key Script: R-Based Software Toolbox for Untargeted Metabolomics with Bladder Cancer Biomarkers Discovery Case Study.

Large-scale untargeted LC-MS-based metabolomic profiling is a valuable source for systems biology and biomarker discovery. Data analysis and processing are major tasks due to the high complexity of generated signals and the presence of unwanted variations. In the present study, we introduce an R-based open-source collection of scripts called OUKS (Omics Untargeted Key Script), which provides comprehensive data processing. OUKS is developed by integrating various R packages and metabolomics software tools and can be easily set up and prepared to create a custom pipeline. Novel computational features are related to quality control samples-based signal processing and are implemented by gradient boosting, tree-based, and other nonlinear regression algorithms. Bladder cancer biomarkers discovery study which is based on untargeted LC-MS profiling of urine samples is performed to demonstrate exhaustive functionality of the developed software tool. Unique examination among dozens of metabolomics-specific data curation methods was carried out at each processing step. As a result, potential biomarkers were identified, statistically validated, and described by metabolism disorders. Our study demonstrates that OUKS helps to make untargeted LC-MS metabolomic profiling with the latest computational features readily accessible in a ready-to-use unified manner to a research community.

Adsorption of water and n-hexane on pristine and oxidized carbon nanotube supports of cobalt-based Fischer-Tropsch catalysts.

Adsorption of model polar (water) and non-polar (n-hexane) compounds on the surface of oxidized and non-oxidized carbon nanotube (CNT) supports at different stages of Co/CNT catalyst preparation has been studied to reveal the influence of the surface functionalization of the CNT support on the catalyst selectivity in Fischer-Tropsch synthesis (FTS). Dynamic vapor sorption experiments showed that defunctionalization of the surface of the CNT support during catalyst annealing and reduction led to its hydrophobization and, as a result, no noticeable difference was observed between the adsorption properties of the oxidized and non-oxidized supports towards water and hydrocarbons. Therefore, oxidation of the CNT support does not significantly affect the adsorption properties of the supported catalyst and it is not a crucial factor for the catalyst selectivity in FTS.