Improvement of quantitative structure-retention relationship models for chromatographic retention prediction of peptides applying individual local partial least squares models

Modeling of small molecule's affinity to phospholipids using IAM-HPLC and QSRR approach enhanced by similarity-based machine algorithms

Immobilized artificial membrane chromatography (IAM) has been proposed as a more biosimilar alternative to classical lipophilicity measurement. Determination of small molecule's affinity to phospholipids can be supported for predicting their behavior in the human body. Therefore, a better understanding of the molecular interaction mechanism between small xenobiotics and phospholipids can accelerate drug discovery. Here, the quantitative structure-retention relationships (QSRR) approach was integrated with mechanistic descriptors calculated using Chemicalize software to propose an easy-to-interpretation QSRR model. Considering the heterogeneous character of the data set, locally weighted least squares kernel regression belonging to similarity-based machine learning methods have been applied. The results showed that lipophilicity, charge, and maximum projection area determine molecule binding to phospholipids. Full validation of the obtained model based on OECD recommendations has been performed and the applicability domain was defined using the probability-oriented distance-based approach. The high values of predictive squared correlation coefficient (Q2), and small root mean square error of prediction (RMSEP), 0.812 and 6.739, respectively, confirmed that the obtained QSRR model is not well-fitted to the training data but also showed prediction power. Additionally, only 1.5% of molecules from the training set and 2.8% from the validation test are outside the applicability domain, confirming great predictive abilities.

A practical strategy enabling more reliable identification of ginsenosides from Panax quinquefolius flower by dimension-enhanced liquid chromatography/mass spectrometry and quantitative structure-retention relationship-based retention behavior prediction

To accurately identify the metabolites is crucial in a number of research fields, and discovery of new compounds from the natural products can benefit the development of new drugs. However, the preferable phytochemistry or liquid chromatography/mass spectrometry approach is time-/labor-extensive or receives unconvincing identifications. Herein, we presented a strategy, by integrating offline two-dimensional liquid chromatography/ion mobility-quadrupole time-of-flight mass spectrometry (2D-LC/IM-QTOF-MS), exclusion list-containing high-definition data-dependent acquisition (HDDDA-EL), and quantitative structure-retention relationship (QSRR) prediction of the retention time (tR), to facilitate the in-depth and more reliable identification of herbal components and thus to discover new compounds more efficiently. Using the saponins in Panax quinquefolius flower (PQF) as a case, high orthogonality (0.79) in separating ginsenosides was enabled by configuring the XBridge Amide and CSH C18 columns. HDDDA-EL could improve the coverage in MS2 acquisition by 2.26 folds compared with HDDDA (2933 VS 1298). Utilizing 106 reference compounds, an accurate QSRR prediction model (R2 = 0.9985 for the training set and R2 = 0.88 for the validation set) was developed based on Gradient Boosting Machine (GBM), by which the predicted tR matching could significantly reduce the isomeric candidates identification for unknown ginsenosides. Isolation and establishment of the structures of two malonylginsenosides by NMR partially verified the practicability of the integral strategy. By these efforts, 421 ginsenosides were identified or tentatively characterized, and 284 thereof were not ever reported from the Panax species. The current strategy is thus powerful in the comprehensive metabolites characterization and rapid discovery of new compounds from the natural products.

Annotation of Dipeptides and Tripeptides Derivatized via Dansylation Based on Liquid Chromatography-Mass Spectrometry and Iterative Quantitative Structure Retention Relationship

Small peptides such as dipeptides and tripeptides show various biological activities in organisms. However, methods for identifying dipeptides/tripeptides from complex biological samples are lacking. Here, an annotation strategy involving the derivatization of dipeptides and tripeptides via dansylation was suggested based on liquid chromatography-mass spectrometry (LC-MS) and iterative quantitative structure retention relationship (QSRR) to choose dipeptides/tripeptides by using a small number of standards. First, the LC-autoMS/MS method and initial QSRR model were built based on 25 selected grid-dipeptides and 18 test-dipeptides. To achieve high-coverage detection, dipeptide/tripeptide pools containing abundant dipeptides/tripeptides were then obtained from four dansylated biological samples including serum, tissue, feces, and soybean paste by using the parameter-optimized LC-autoMS/MS method. The QSRR model was further optimized through an iterative train-by-pick strategy. Based on the specific fragments and t_R tolerances, 198 dipeptides and 149 tripeptides were annotated. The dipeptides at lower annotation levels were verified by using authentic standards and grid-correlation analysis. Finally, variation in serum dipeptides/tripeptides of three different liver diseases including hepatitis B infection, liver cirrhosis, and hepatocellular carcinoma was characterized. Dipeptides with N-prolinyl, C-proline, N-glutamyl, and N-valinyl generally increased with disease severity. In conclusion, this study provides an efficient strategy for annotating dipeptides/tripeptides from complex samples.