Author Archives: Vera

The work employed a software tool automating data analysis stages for LC-MS (High Resolution) data. This included selecting chromatographic peaks related to the compound, retrieving mass spectral information, assigning potential structures through theoretical fragmentation comparison with experimental m/z values, and scoring solutions based on fragment analysis. Results from different experimental conditions are clustered into a unified experiment entity and stored in a database. For each compound two distinct algorithms have been employed for peak selection, allowing for outcome comparison. After data consolidation, manual interpretation is performed according to predefined criteria. Data from different acquisition modes has been processed, and two structure visualization methods are presented: an expanded form depicting all atoms and bonds, and a non-expanded form linking monomer acronyms.

The aim of this study is to describe new algorithms/approaches for automated LC-MS (High Resolution) data analysis that addresses the mentioned challenges encountered in the processing of macromolecules. These challenges encompass optimizing the input and visualization of chemical structures and degradation products. Additionally, it has successfully optimized the reduction of processing memory and time consumption (from 2 hours to 25 minutes) in the execution of algorithms for potential structure generation and fragmentation. Furthermore, the proposed methods aim to provide a workflow capable of interpreting results across various data acquisition formats and modes.

Analysis was conducted on six datasets spanning a molecular range of 700 to 15,000 Da. These datasets consist of both linear and cyclic peptides, incorporating natural and unnatural amino acids, as well as an oligonucleotide. Specifically, dataset-1 comprises nine commercially available peptides, dataset-2 includes one commercially available peptide and four synthetic analogues, dataset-3 involves a natural peptide hormone and seven synthetic analogues, dataset-4 features an antisense oligonucleotide, dataset-5 contains 28 commercially available peptides, and dataset-6 is composed of a peptide hormone.

Comparisons of the results obtained for certain compounds with those of prior studies have enabled a comprehensive evaluation across various parameters. This evaluation encompasses aspects such as the number and structure of identified metabolites, along with a consideration of the time consumed during the data processing step.

The results obtained indicate that, in larger molecules, the most abundant mass algorithm demonstrated higher scores and a greater number of matches, and therefore greater confidence in the accurate prediction of metabolite structures. Furthermore, this study shows three visualization options for representing macromolecules during data analysis. This visualization algorithm allows the combination of monomer and atom/bond notation, facilitating a clear depiction of metabolic changes in the molecular structure.

Metabolite identification data was gathered from public repositories and processed using specialized software to automatically identify corresponding metabolites. A field expert later examined the results, manually categorizing them as either true or false positives.

Extracting features from each metabolite peak involved utilizing chromatographic peak data, mass spectrometry data, and kinetic data where applicable. The dataset was then divided into training and test sets, with an 80/20 split. Within the training set, machine learning models were developed using gradient boosting decision trees (GBDTs). Hyperparameters underwent tuning through random search and 5-fold cross-validation. The best-performing model was evaluated using the test set. This entire process was repeated for each of the two distinct experimental protocols present in the data.

Two distinct experimental protocols were present in the data. The primary difference between them lay in the number of incubation time points: one protocol involved 5 time points while the other featured just 1. Consequently, a separate model was developed for each protocol, given that the analysis of the data varied significantly, particularly with the former protocol requiring the incorporation of kinetic data.

For experiments following the 1-time-point protocol, a total of 2,570 metabolite peaks were examined, with 1,703 determined as false positives (66.26%). The primary objective of these experiments was the identification of gluthatione conjugates, formed by appending a gluthatione moiety to a parent molecule. Data for testing was constructed by splitting experiments, resulting in 431 metabolites, of which 287 were false positives (66.60%). The best-performing model achieved a recall of 93.73% and a precision of 89.67% on the test set.

Experiments adhering to the 5-time-point protocol involved a total of 1,543 metabolite peaks, with 1,068 identified as false positives (69.22%). The primary focus of these experiments was soft spot identification. Test data was generated by experiment splits, yielding 419 metabolites, of which 287 were false positives (68.50%). The top-performing model attained a recall of 97.56% and a precision of 93.33% on the test set.

Beyond evaluating the predictive performance of the models, the SHapley Additive exPlanations (SHAP) method was employed to analyze feature importances. The most significant features for each prediction were extracted and aggregated by category or source. This facilitated a clear understanding of the main reasons behind predictions, which were empirically verified to be accurate in most cases. Consequently, this method could serve as a guiding tool to aid experts in the manual inspection of metabolite identification results, as well as in reassessing and rectifying previous manual annotations.