Dataset Information

Drug Repurposing Prediction for Immune-Mediated Cutaneous Diseases using a Word-Embedding-Based Machine Learning Approach.

ABSTRACT: Immune-mediated diseases affect more than 20% of the population, and many autoimmune diseases affect the skin. Drug repurposing (or repositioning) is a cost-effective approach for finding drugs that can be used to treat diseases for which they are currently not prescribed. We implemented an efficient bioinformatics approach using word embedding to summarize drug information from more than 20 million articles and applied machine learning to model the drug-disease relationship. We trained our drug repurposing approach separately on nine cutaneous diseases (including psoriasis, atopic dermatitis, and alopecia areata) and eight other immune-mediated diseases and obtained a mean area under the receiver operating characteristic of 0.93 in cross-validation. Focusing in particular on psoriasis, a chronic inflammatory condition of skin that affects more than 100 million people worldwide, we were able to confirm drugs that are known to be effective for psoriasis and to identify potential candidates used to treat other diseases. Furthermore, the targets of drug candidates predicted by our approach were significantly enriched among genes differentially expressed in psoriatic lesional skin from a large-scale RNA sequencing cohort. Although our algorithm cannot be used to determine clinical efficacy, our work provides an approach for suggesting drugs for repurposing to immune-mediated cutaneous diseases.

SUBMITTER: Patrick MT

PROVIDER: S-EPMC6387843 | biostudies-literature | 2019 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Drug Repurposing Prediction for Immune-Mediated Cutaneous Diseases using a Word-Embedding-Based Machine Learning Approach.

Patrick Matthew T MT Raja Kalpana K Miller Keylonnie K Sotzen Jason J Gudjonsson Johann E JE Elder James T JT Tsoi Lam C LC

The Journal of investigative dermatology 20181017 3

Immune-mediated diseases affect more than 20% of the population, and many autoimmune diseases affect the skin. Drug repurposing (or repositioning) is a cost-effective approach for finding drugs that can be used to treat diseases for which they are currently not prescribed. We implemented an efficient bioinformatics approach using word embedding to summarize drug information from more than 20 million articles and applied machine learning to model the drug-disease relationship. We trained our drug ...[more]

PMID: 30342048

Similar Datasets

Project description:Drug repurposing has broad importance in planetary health for therapeutics innovation in infectious diseases as well as common or rare chronic human diseases. Drug repurposing has also proved important to develop interventions against the COVID-19 pandemic. We propose a new approach for drug repurposing involving two-stage prediction and machine learning. First, diseases are clustered by gene expression on the premise that similar patterns of altered gene expression imply critical pathways shared in different disease conditions. Next, drug efficacy is assessed by the reversibility of abnormal gene expression, and results are clustered to identify repurposing targets. To cluster similar diseases, gene expression data from 262 cases of 31 diseases and 268 controls were analyzed by Uniform Manifold Approximation and Projection for Dimension Reduction followed by k-means to optimize the number of clusters. For evaluation, we examined disease-specific gene expression data for inclusion, body myositis, polymyositis, and dermatomyositis (DM), and used LINCS L1000 characteristic direction signatures search engine (L1000CDS2) to obtain lists of small-molecule compounds that reversed the expression patterns of these specifically altered genes as candidates for drug repurposing. Finally, the functions of affected genes were analyzed by Gene Set Enrichment Analysis to examine consistency with expected drug efficacy. Consequently, we found disease-specific gene expression, and importantly, identified 20 drugs such as BMS-387032, phorbol-12-myristate-13-acetate, mitoxantrone, alvocidib, and vorinostat as candidates for repurposing. These were previously noted to be effective against two of the three diseases, and have a high probability of being effective against the other. That is, inclusion body myositis and DM. The two-stage prediction approach to drug repurposing presented here offers innovation to inform future drug discovery and clinical trials in a variety of human diseases.

Project description:BackgroundThe exponential growth of location-based social media (LBSM) data has ushered in novel prospects for investigating the urban food environment in health geography research. However, previous studies have primarily relied on word dictionaries with a limited number of food words and employed common-sense categorizations to determine the healthiness of those words. To enhance the analysis of the urban food environment using LBSM data, it is crucial to develop a more comprehensive list of food-related words. Within the context, this study delves into the exploration of expanding food-related words along with their associated energy densities.MethodsThis study addresses the aforementioned research gap by introducing a novel methodology for expanding the food-related word dictionary and predicting energy densities. Seed words are generated from official and crowdsourced food composition databases, and new food words are discovered by clustering food words within the word embedding space using the Gaussian mixture model. Machine learning models are employed to predict the energy density classifications of these food words based on their feature vectors. To ensure a thorough exploration of the prediction problem, ten widely used machine learning models are evaluated.ResultsThe approach successfully expands the food-related word dictionary and accurately predicts food energy density (reaching 91.62%.). Through a comparison of the newly expanded dictionary with the initial seed words and an analysis of Yelp reviews in the city of Toronto, we observe significant improvements in identifying food words and gaining a deeper understanding of the food environment.ConclusionsThis study proposes a novel method to expand food-related vocabulary and predict the food energy density based on machine learning and word embedding. This method makes a valuable contribution to building a more comprehensive list of food words that can be used in geography and public health studies by mining geotagged social media data.

Project description:The identification of drug/compound-target interactions (DTIs) constitutes the basis of drug discovery, for which computational predictive approaches have been developed. As a relatively new data-driven paradigm, proteochemometric (PCM) modeling utilizes both protein and compound properties as a pair at the input level and processes them via statistical/machine learning. The representation of input samples (i.e., proteins and their ligands) in the form of quantitative feature vectors is crucial for the extraction of interaction-related properties during the artificial learning and subsequent prediction of DTIs. Lately, the representation learning approach, in which input samples are automatically featurized via training and applying a machine/deep learning model, has been utilized in biomedical sciences. In this study, we performed a comprehensive investigation of different computational approaches/techniques for protein featurization (including both conventional approaches and the novel learned embeddings), data preparation and exploration, machine learning-based modeling, and performance evaluation with the aim of achieving better data representations and more successful learning in DTI prediction. For this, we first constructed realistic and challenging benchmark datasets on small, medium, and large scales to be used as reliable gold standards for specific DTI modeling tasks. We developed and applied a network analysis-based splitting strategy to divide datasets into structurally different training and test folds. Using these datasets together with various featurization methods, we trained and tested DTI prediction models and evaluated their performance from different angles. Our main findings can be summarized under 3 items: (i) random splitting of datasets into train and test folds leads to near-complete data memorization and produce highly over-optimistic results, as a result, should be avoided, (ii) learned protein sequence embeddings work well in DTI prediction and offer high potential, despite interaction-related properties (e.g., structures) of proteins are unused during their self-supervised model training, and (iii) during the learning process, PCM models tend to rely heavily on compound features while partially ignoring protein features, primarily due to the inherent bias in DTI data, indicating the requirement for new and unbiased datasets. We hope this study will aid researchers in designing robust and high-performing data-driven DTI prediction systems that have real-world translational value in drug discovery.

Dataset Information

Drug Repurposing Prediction for Immune-Mediated Cutaneous Diseases using a Word-Embedding-Based Machine Learning Approach.

Publications

Drug Repurposing Prediction for Immune-Mediated Cutaneous Diseases using a Word-Embedding-Based Machine Learning Approach.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets