Development and Validation of an Explainable Machine Learning-Based Prediction Model for Drug-Food Interactions from Chemical Structures

Possible drug-food constituent interactions (DFIs) could change the intended efficiency of particular therapeutics in medical practice. The increasing number of multiple-drug prescriptions leads to the rise of drug-drug interactions (DDIs) and DFIs. These adverse interactions lead to other implications, e.g., the decline in medicament's effect, the withdrawals of various medications, and harmful impacts on the patients' health. However, the importance of DFIs remains underestimated, as the number of studies on these topics is constrained. Recently, scientists have applied artificial intelligence-based models to study DFIs. However, there were still some limitations in data mining, input, and detailed annotations. This study proposed a novel prediction model to address the limitations of previous studies. In detail, we extracted 70,477 food compounds from the FooDB database and 13,580 drugs from the DrugBank database. We extracted 3780 features from each drug-food compound pair. The optimal model was eXtreme Gradient Boosting (XGBoost). We also validated the performance of our model on one external test set from a previous study which contained 1922 DFIs. Finally, we applied our model to recommend whether a drug should or should not be taken with some food compounds based on their interactions. The model can provide highly accurate and clinically relevant recommendations, especially for DFIs that may cause severe adverse events and even death. Our proposed model can contribute to developing more robust predictive models to help patients, under the supervision and consultants of physicians, avoid DFI adverse effects in combining drugs and foods for therapy.

Chemputation and the Standardization of Chemical Informatics

The explosion in the use of machine learning for automated chemical reaction optimization is gathering pace. However, the lack of a standard architecture that connects the concept of chemical transformations universally to software and hardware provides a barrier to using the results of these optimizations and could cause the loss of relevant data and prevent reactions from being reproducible or unexpected findings verifiable or explainable. In this Perspective, we describe how the development of the field of digital chemistry or chemputation, that is the universal code-enabled control of chemical reactions using a standard language and ontology, will remove these barriers allowing users to focus on the chemistry and plug in algorithms according to the problem space to be explored or unit function to be optimized. We describe a standard hardware (the chemical processing programming architecture-the ChemPU) to encompass all chemical synthesis, an approach which unifies all chemistry automation strategies, from solid-phase peptide synthesis, to HTE flow chemistry platforms, while at the same time establishing a publication standard so that researchers can exchange chemical code (χDL) to ensure reproducibility and interoperability. Not only can a vast range of different chemistries be plugged into the hardware, but the ever-expanding developments in software and algorithms can also be accommodated. These technologies, when combined will allow chemistry, or chemputation, to follow computation-that is the running of code across many different types of capable hardware to get the same result every time with a low error rate.

Predicting the Mixing Behavior of Aqueous Solutions Using a Machine Learning Framework

The most direct approach to determining if two aqueous solutions will phase-separate upon mixing is to exhaustively screen them in a pair-wise fashion. This is a time-consuming process that involves preparation of numerous stock solutions, precise transfer of highly concentrated and often viscous solutions, exhaustive agitation to ensure thorough mixing, and time-sensitive monitoring to observe the presence of emulsion characteristics indicative of phase separation. Here, we examined the pair-wise mixing behavior of 68 water-soluble compounds by observing the formation of microscopic phase boundaries and droplets of 2278 unique 2-component solutions. A series of machine learning classifiers (artificial neural network, random forest, k-nearest neighbors, and support vector classifier) were then trained on physicochemical property data associated with the 68 compounds and used to predict their miscibility upon mixing. Miscibility predictions were then compared to the experimental observations. The random forest classifier was the most successful classifier of those tested, displaying an average receiver operator characteristic area under the curve of 0.74. The random forest classifier was validated by removing either one or two compounds from the input data, training the classifier on the remaining data and then predicting the miscibility of solutions involving the removed compound(s) using the classifier. The accuracy, specificity, and sensitivity of the random forest classifier were 0.74, 0.80, and 0.51, respectively, when one of the two compounds to be examined was not represented in the training data. When asked to predict the miscibility of two compounds, neither of which were represented in the training data, the accuracy, specificity, and sensitivity values for the random forest classifier were 0.70, 0.82 and 0.29, respectively. Thus, there is potential for this machine learning approach to improve the design of screening experiments to accelerate the discovery of aqueous two-phase systems for numerous scientific and industrial applications.

Systematic Chemogenetic Library Assembly

Chemogenetic libraries, collections of well-defined chemical probes, provide tremendous value to biomedical research but require substantial effort to ensure diversity as well as quality of the contents. We have assembled a chemogenetic library by data mining and crowdsourcing institutional expertise. We are sharing our approach, lessons learned, and disclosing our current collection of 4,185 compounds with their primary annotated gene targets (https://github.com/Novartis/MoaBox). This physical collection is regularly updated and used broadly both within Novartis and in collaboration with external partners.

Exploring the Regulation Mechanism of Xihuang Pill, Olibanum and β-Boswellic Acid on the Biomolecular Network of Triple-Negative Breast Cancer Based on Transcriptomics and Chemical Informatics Methodology

Background

Xihuang Pill (XHP) is mainly used to treat “Ru Yan (breast cancer)”. Evidence-based medical evidence and showed that XHP improves the efficacy of chemotherapy and reduced chemotherapy-induced toxicity in breast cancer patients. However, the mechanism of XHP against breast cancer is not clear.

Methods

The effect of XHP extract on cell half-inhibitory concentration (IC50) and cell viability of MD-MB-231 cells was detected by CCK-8 method. The cell inhibition rate of MDA-MB-453 cells were detected by MTT method. Apoptosis was detected by flow cytometry, cell transfer ability was detected by Transwell method, and cell proliferation ability was detected by colony formation assay. The expression of Notch1, β-catenin and c-myc mRNA in MDA-MB-453 cells were detected by real-time fluorescence quantitative PCR. Then, chemical informatics and transcriptomics methodology was utilized to predict the potential compounds and targets of XHP, and collect triple negative breast cancer (TNBC) genes and the data of Olibanum and β-boswellic acid intervention MD-MB-231 cells (from GSE102891). The cytoscape software was utilized to undergo network construction and network analysis. Finally, the data from the network analysis was imported into the DAVID database for enrichment analysis of signaling pathways and biological processes.

Results

The IC50 was 15.08 g/L (for MD-MB-231 cells). After interfering with MD-MB-231 cells with 15.08 g/L XHP extract for 72 h, compared with the control group, the cell viability, migration and proliferation was significantly decreased, while early apoptosis and late apoptosis were significantly increased (P < 0.01). After interfering with MDA-MB-453 cells with 6 g/L XHP extract for 72 h, compared with the control group, the cell inhibition and apoptosis rate increased, while the expression of Notch1, β-catenin and c-myc mRNA decreased. (P < 0.05). The chemical informatics and transcriptomics analysis showed that four networks were constructed and analyzed: (1) potential compounds-potential targets network of XHP; (2) XHP-TNBC PPI network; (3) DEGs PPI network of Olibanum-treated MD-MB 231 cells; (4) DEGs PPI network of β-boswellic acid -treated MD-MB 231 cells. Several anti-TNBC biological processes, signaling pathways, targets and so on were obtained.

Conclusion

XHP may exert anti-TNBC effects through regulating biological processes, signaling pathways, targets found in this study.

Structure Activity Relationships (SARs) Using a Structurally Diverse Drug Database: Validating Success of Predictor Tools

ADME/Tox (absorption, distribution, metabolism, elimination and toxicity) technology is traditionally associated as a tool in the drug discovery process which is often used to predict the efficiency of drug adsorption, distribution, metabolic pathways, and elimination. For the past four years we have been involved in an effort to evaluate readily available Food and Drug Administration (FDA) consumer drug profiles and pharmacological data. Portable Document Format (PDF) data from drug profiles available on the FDA Drug Information website were used to create a searchable FDA Consumer Drug Database^© using Bio-Rad's KnowItAll^® platform which includes ADME/Tox in silico predictors. 14 pertinent pharmaceutical and pharmacological properties were collected for 75 structurally diverse consumer prescription drugs, and for several drugs, not all properties were completely populated. The major objective of this investigation was to validate the platforms prediction models for plasma protein binding (PPB) and bioavailability (BIO).

Extracting Relevant Information from FDA Drug Files to Create a Structurally Diverse Drug Database Using KnowItAll®

Each Food and Drug Administration (FDA) consumer drug information file contains an inordinate amount of useful chemical, pharmaceutical, and pharmacological data. These files profile approved drugs by chemical structure, solubility, absorption, distribution, metabolism, elimination, toxicity (ADME/Tox), and possible adverse reactions. The ability to utilize this data in the classroom is a new approach to connect theory, technology, and reality. The KnowItAll^® Informatics System available through Bio-Rad Laboratories, Philadelphia, PA, offers fully integrated software and/or database desktop solutions. It holds a large collection of in silico ADME/Tox predictors and is a chemical informatics platform used to record experimental data. This project had three goals: (1) extract relevant information for 75 drugs from their freely available FDA drug files (limited to orally administrated drugs, pro-drugs, having a chemical structure), (2) build a database so this extracted FDA information is indexed for search and analysis, and when completed, (3) undergraduates involved in such a project should be capable of harvesting useful chemical, pharmaceutical, and pharmacological information; be adept in computational chemistry software tools; and should gain an enhanced vocabulary and new insights into organic chemistry, molecular biology, and physiology.