An end-to-end deep learning framework for translating mass spectra to de-novo molecules.

Elucidating the structure of a chemical compound is a fundamental task in chemistry with applications in multiple domains including drug discovery, precision medicine, and biomarker discovery. The common practice for elucidating the structure of a compound is to obtain a mass spectrum and subsequently retrieve its structure from spectral databases. However, these methods fail for novel molecules that are not present in the reference database. We propose Spec2Mol, a deep learning architecture for molecular structure recommendation given mass spectra alone. Spec2Mol is inspired by the Speech2Text deep learning architectures for translating audio signals into text. Our approach is based on an encoder-decoder architecture. The encoder learns the spectra embeddings, while the decoder, pre-trained on a massive dataset of chemical structures for translating between different molecular representations, reconstructs SMILES sequences of the recommended chemical structures. We have evaluated Spec2Mol by assessing the molecular similarity between the recommended structures and the original structure. Our analysis showed that Spec2Mol is able to identify the presence of key molecular substructures from its mass spectrum, and shows on par performance, when compared to existing fragmentation tree methods particularly when test structure information is not available during training or present in the reference database.

Machine Learning-Guided Three-Dimensional Printing of Tissue Engineering Scaffolds.

Various material compositions have been successfully used in 3D printing with promising applications as scaffolds in tissue engineering. However, identifying suitable printing conditions for new materials requires extensive experimentation in a time and resource-demanding process. This study investigates the use of Machine Learning (ML) for distinguishing between printing configurations that are likely to result in low-quality prints and printing configurations that are more promising as a first step toward the development of a recommendation system for identifying suitable printing conditions. The ML-based framework takes as input the printing conditions regarding the material composition and the printing parameters and predicts the quality of the resulting print as either "low" or "high." We investigate two ML-based approaches: a direct classification-based approach that trains a classifier to distinguish between low- and high-quality prints and an indirect approach that uses a regression ML model that approximates the values of a printing quality metric. Both modes are built upon Random Forests. We trained and evaluated the models on a dataset that was generated in a previous study, which investigated fabrication of porous polymer scaffolds by means of extrusion-based 3D printing with a full-factorial design. Our results show that both models were able to correctly label the majority of the tested configurations while a simpler linear ML model was not effective. Additionally, our analysis showed that a full factorial design for data collection can lead to redundancies in the data, in the context of ML, and we propose a more efficient data collection strategy.

Prediction of drug metabolites using neural machine translation.

Metabolic processes in the human body can alter the structure of a drug affecting its efficacy and safety. As a result, the investigation of the metabolic fate of a candidate drug is an essential part of drug design studies. Computational approaches have been developed for the prediction of possible drug metabolites in an effort to assist the traditional and resource-demanding experimental route. Current methodologies are based upon metabolic transformation rules, which are tied to specific enzyme families and therefore lack generalization, and additionally may involve manual work from experts limiting scalability. We present a rule-free, end-to-end learning-based method for predicting possible human metabolites of small molecules including drugs. The metabolite prediction task is approached as a sequence translation problem with chemical compounds represented using the SMILES notation. We perform transfer learning on a deep learning transformer model for sequence translation, originally trained on chemical reaction data, to predict the outcome of human metabolic reactions. We further build an ensemble model to account for multiple and diverse metabolites. Extensive evaluation reveals that the proposed method generalizes well to different enzyme families, as it can correctly predict metabolites through phase I and phase II drug metabolism as well as other enzymes. Compared to existing rule-based approaches, our method has equivalent performance on the major enzyme families while it additionally finds metabolites through less common enzymes. Our results indicate that the proposed approach can provide a comprehensive study of drug metabolism that does not restrict to the major enzyme families and does not require the extraction of transformation rules.

Machine Learning Guided Atom Mapping of Metabolic Reactions.

Atom mapping of a chemical reaction is a mapping between the atoms in the reactant molecules and the atoms in the product molecules. It encodes the underlying reaction mechanism and, as such, constitutes essential information in computational studies in drug design. Various techniques have been investigated for the automatic computation of the atom mapping of a chemical reaction, approaching the problem as a graph matching problem. The graph abstraction of the chemical problem, though, eliminates crucial chemical information. There have been efforts for enhancing the graph representation by introducing the bond stabilities as edge weights, as they are estimated based on experimental evidence. Here, we present a fully automated optimization-based approach, named AMLGAM (Automated Machine Learning Guided Atom Mapping), that uses machine learning techniques for the estimation of the bond stabilities based on the chemical environment of each bond. The optimization method finds the reaction mechanism which favors the breakage/formation of the less stable bonds. We evaluated our method on a manually curated data set of 382 chemical reactions and ran our method on a much larger and diverse data set of 7400 chemical reactions. We show that the proposed method improves the accuracy over existing techniques based on results published by earlier studies on a common data set and is capable of handling unbalanced reactions.

A Review of Emerging Technologies for the Management of Diabetes Mellitus.

OBJECTIVE: High prevalence of diabetes mellitus (DM) along with the poor health outcomes and the escalated costs of treatment and care poses the need to focus on prevention, early detection and improved management of the disease. The aim of this paper is to present and discuss the latest accomplishments in sensors for glucose and lifestyle monitoring along with clinical decision support systems (CDSSs) facilitating self-disease management and supporting healthcare professionals in decision making. METHODS: A critical literature review analysis is conducted focusing on advances in: 1) sensors for physiological and lifestyle monitoring, 2) models and molecular biomarkers for predicting the onset and assessing the progress of DM, and 3) modeling and control methods for regulating glucose levels. RESULTS: Glucose and lifestyle sensing technologies are continuously evolving with current research focusing on the development of noninvasive sensors for accurate glucose monitoring. A wide range of modeling, classification, clustering, and control approaches have been deployed for the development of the CDSS for diabetes management. Sophisticated multiscale, multilevel modeling frameworks taking into account information from behavioral down to molecular level are necessary to reveal correlations and patterns indicating the onset and evolution of DM. CONCLUSION: Integration of data originating from sensor-based systems and electronic health records combined with smart data analytics methods and powerful user centered approaches enable the shift toward preventive, predictive, personalized, and participatory diabetes care. SIGNIFICANCE: The potential of sensing and predictive modeling approaches toward improving diabetes management is highlighted and related challenges are identified.