Predicting biological pathways of chemical compounds with a profile-inspired approach

Heterogeneous network approaches to protein pathway prediction

Understanding protein-protein interactions (PPIs) and the pathways they comprise is essential for comprehending cellular functions and their links to specific phenotypes. Despite the prevalence of molecular data generated by high-throughput sequencing technologies, a significant gap remains in translating this data into functional information regarding the series of interactions that underlie phenotypic differences. In this review, we present an in-depth analysis of heterogeneous network methodologies for modeling protein pathways, highlighting the critical role of integrating multifaceted biological data. It outlines the process of constructing these networks, from data representation to machine learning-driven predictions and evaluations. The work underscores the potential of heterogeneous networks in capturing the complexity of proteomic interactions, thereby offering enhanced accuracy in pathway prediction. This approach not only deepens our understanding of cellular processes but also opens up new possibilities in disease treatment and drug discovery by leveraging the predictive power of comprehensive proteomic data analysis.

Identification of Key Toxic Substances Considering Metabolic Activation: A Combination of Transcriptome and Nontarget Analysis

There have been numerous studies using effect-directed analysis (EDA) to identify key toxic substances present in source and drinking water, but none of these studies have considered the effects of metabolic activation. This study developed a comprehensive method including a pretreatment process based on an in vitro metabolic activation system, a comprehensive biological effect evaluation based on concentration-dependent transcriptome (CDT), and a chemical feature identification based on nontarget chemical analysis (NTA), to evaluate the changes in the toxic effects and differences in the chemical composition after metabolism. Models for matching metabolites and precursors as well as data-driven identification methods were further constructed to identify toxic metabolites and key toxic precursor substances in drinking water samples from the Yangtze River. After metabolism, the metabolic samples showed a general trend of reduced toxicity in terms of overall biological potency (mean: 3.2-fold). However, metabolic activation led to an increase in some types of toxic effects, including pathways such as excision repair, mismatch repair, protein processing in endoplasmic reticulum, nucleotide excision repair, and DNA replication. Meanwhile, metabolic samples showed a decrease (17.8%) in the number of peaks and average peak area after metabolism, while overall polarity, hydrophilicity, and average molecular weight increased slightly (10.3%). Based on the models for matching of metabolites and precursors and the data-driven identification methods, 32 chemicals were efficiently identified as key toxic substances as main contributors to explain the different transcriptome biological effects such as cellular component, development, and DNA damage related, including 15 industrial compounds, 7 PPCPs, 6 pesticides, and 4 natural products. This study avoids the process of structure elucidation of toxic metabolites and can trace them directly to the precursors based on MS spectra, providing a new idea for the identification of key toxic pollutants of metabolites.

Benchmark Dataset for Training Machine Learning Models to Predict the Pathway Involvement of Metabolites

Metabolic pathways are a human-defined grouping of life sustaining biochemical reactions, metabolites being both the reactants and products of these reactions. But many public datasets include identified metabolites whose pathway involvement is unknown, hindering metabolic interpretation. To address these shortcomings, various machine learning models, including those trained on data from the Kyoto Encyclopedia of Genes and Genomes (KEGG), have been developed to predict the pathway involvement of metabolites based on their chemical descriptions; however, these prior models are based on old metabolite KEGG-based datasets, including one benchmark dataset that is invalid due to the presence of over 1500 duplicate entries. Therefore, we have developed a new benchmark dataset derived from the KEGG following optimal standards of scientific computational reproducibility and including all source code needed to update the benchmark dataset as KEGG changes. We have used this new benchmark dataset with our atom coloring methodology to develop and compare the performance of Random Forest, XGBoost, and multilayer perceptron with autoencoder models generated from our new benchmark dataset. Best overall weighted average performance across 1000 unique folds was an F1 score of 0.8180 and a Matthews correlation coefficient of 0.7933, which was provided by XGBoost binary classification models for 11 KEGG-defined pathway categories.

Benchmark dataset for training machine learning models to predict the pathway involvement of metabolites

Metabolic pathways are a human-defined grouping of life sustaining biochemical reactions, metabolites being both the reactants and products of these reactions. But many public datasets include identified metabolites whose pathway involvement is unknown, hindering metabolic interpretation. To address these shortcomings, various machine learning models, including those trained on data from the Kyoto Encyclopedia of Genes and Genomes (KEGG), have been developed to predict the pathway involvement of metabolites based on their chemical descriptions; however, these prior models are based on old metabolite KEGG-based datasets, including one benchmark dataset that is invalid due to the presence of over 1500 duplicate entries. Therefore, we have developed a new benchmark dataset derived from the KEGG following optimal standards of scientific computational reproducibility and including all source code needed to update the benchmark dataset as KEGG changes. We have used this new benchmark dataset with our atom coloring methodology to develop and compare the performance of Random Forest, XGBoost, and multilayer perceptron with autoencoder models generated from our new benchmark dataset. Best overall weighted average performance across 1000 unique folds was an F1-score of 0.8180 and Matthews correlation coefficient of 0.7933, which was provided by XGBoost binary classification models for 11 KEGG-defined pathway categories.

Sensbio: an online server for biosensor design

Allosteric transcription factor (aTF) based biosensors can be used to engineer genetic circuits for a wide range of applications. The literature and online databases contain hundreds of experimentally validated molecule-TF pairs; however, the knowledge is scattered and often incomplete. Additionally, compared to the number of compounds that can be produced in living systems, those with known associated TF-compound interactions are low. For these reasons, new tools that help researchers find new possible TF-ligand pairs are called for. In this work, we present Sensbio, a computational tool that through similarity comparison against a TF-ligand reference database, is able to identify putative transcription factors that can be activated by a given input molecule. In addition to the collection of algorithms, an online application has also been developed, together with a predictive model created to find new possible matches based on machine learning.

Predicting protein network topology clusters from chemical structure using deep learning

Comparing chemical structures to infer protein targets and functions is a common approach, but basing comparisons on chemical similarity alone can be misleading. Here we present a methodology for predicting target protein clusters using deep neural networks. The model is trained on clusters of compounds based on similarities calculated from combined compound-protein and protein-protein interaction data using a network topology approach. We compare several deep learning architectures including both convolutional and recurrent neural networks. The best performing method, the recurrent neural network architecture MolPMoFiT, achieved an F1 score approaching 0.9 on a held-out test set of 8907 compounds. In addition, in-depth analysis on a set of eleven well-studied chemical compounds with known functions showed that predictions were justifiable for all but one of the chemicals. Four of the compounds, similar in their molecular structure but with dissimilarities in their function, revealed advantages of our method compared to using chemical similarity.