Drug Therapeutic-Use Class Prediction and Repurposing Using Graph Convolutional Networks

An important stage in the process of discovering new drugs is when candidate molecules are tested of their efficacy. It is reported that testing drug efficacy empirically costs billions of dollars in the drug discovery pipeline. As a mechanism of expediting this process, researchers have resorted to using computational methods to predict the action of molecules in silico. Here, we present a way of predicting the therapeutic-use class of drugs from chemical structures only using graph convolutional networks. In comparison with existing methods which use fingerprints or images as training samples, our approach has yielded better results in all metrics under consideration. In particular, validation accuracy increased from 83–88% to 86–90% for single label tasks. Similarly, the model achieved an accuracy of over 88% on new test data. Finally, our multi-label classification model made new predictions which indicated that some of the drugs could have other therapeutic uses other than those indicated in the dataset. We performed a literature-based evaluation of these predictions and found evidence that validates them. This renders the model a potential tool to be used in search of drugs that are candidates for repurposing.

Download Full-text

A deep learning architecture for metabolic pathway prediction

Bioinformatics ◽

10.1093/bioinformatics/btz954 ◽

2019 ◽

Vol 36 (8) ◽

pp. 2547-2553 ◽

Cited By ~ 9

Author(s):

Mayank Baranwal ◽

Abram Magner ◽

Paolo Elvati ◽

Jacob Saldinger ◽

Angela Violi ◽

...

Keyword(s):

Machine Learning ◽

Metabolic Pathway ◽

Molecular Shape ◽

Supplementary Information ◽

Shape Features ◽

Global Features ◽

Class Prediction ◽

Logistic Regression Models ◽

Convolutional Networks ◽

Chemical Structures

Abstract Motivation Understanding the mechanisms and structural mappings between molecules and pathway classes are critical for design of reaction predictors for synthesizing new molecules. This article studies the problem of prediction of classes of metabolic pathways (series of chemical reactions occurring within a cell) in which a given biochemical compound participates. We apply a hybrid machine learning approach consisting of graph convolutional networks used to extract molecular shape features as input to a random forest classifier. In contrast to previously applied machine learning methods for this problem, our framework automatically extracts relevant shape features directly from input SMILES representations, which are atom-bond specifications of chemical structures composing the molecules. Results Our method is capable of correctly predicting the respective metabolic pathway class of 95.16% of tested compounds, whereas competing methods only achieve an accuracy of 84.92% or less. Furthermore, our framework extends to the task of classification of compounds having mixed membership in multiple pathway classes. Our prediction accuracy for this multi-label task is 97.61%. We analyze the relative importance of various global physicochemical features to the pathway class prediction problem and show that simple linear/logistic regression models can predict the values of these global features from the shape features extracted using our framework. Availability and implementation https://github.com/baranwa2/MetabolicPathwayPrediction. Supplementary information Supplementary data are available at Bioinformatics online.

Download Full-text

Learning chemistry: exploring the suitability of machine learning for the task of structure-based chemical ontology classification

Journal of Cheminformatics ◽

10.1186/s13321-021-00500-8 ◽

2021 ◽

Vol 13 (1) ◽

Cited By ~ 1

Author(s):

Janna Hastings ◽

Martin Glauer ◽

Adel Memariani ◽

Fabian Neuhaus ◽

Till Mossakowski

Keyword(s):

Machine Learning ◽

Logistic Regression ◽

Short Term Memory ◽

Chemical Space ◽

Chemical Data ◽

Learning Approaches ◽

Class Prediction ◽

Chemical Structures ◽

Chemical Ontology ◽

Chemical Ontologies

AbstractChemical data is increasingly openly available in databases such as PubChem, which contains approximately 110 million compound entries as of February 2021. With the availability of data at such scale, the burden has shifted to organisation, analysis and interpretation. Chemical ontologies provide structured classifications of chemical entities that can be used for navigation and filtering of the large chemical space. ChEBI is a prominent example of a chemical ontology, widely used in life science contexts. However, ChEBI is manually maintained and as such cannot easily scale to the full scope of public chemical data. There is a need for tools that are able to automatically classify chemical data into chemical ontologies, which can be framed as a hierarchical multi-class classification problem. In this paper we evaluate machine learning approaches for this task, comparing different learning frameworks including logistic regression, decision trees and long short-term memory artificial neural networks, and different encoding approaches for the chemical structures, including cheminformatics fingerprints and character-based encoding from chemical line notation representations. We find that classical learning approaches such as logistic regression perform well with sets of relatively specific, disjoint chemical classes, while the neural network is able to handle larger sets of overlapping classes but needs more examples per class to learn from, and is not able to make a class prediction for every molecule. Future work will explore hybrid and ensemble approaches, as well as alternative network architectures including neuro-symbolic approaches.

Download Full-text

Estimating Drug Efficacy with a Diet-Induced NASH Model in Chimeric Mice with Humanized Livers

Biomedicines ◽

10.3390/biomedicines9111647 ◽

2021 ◽

Vol 9 (11) ◽

pp. 1647

Author(s):

Keishi Kisoh ◽

Go Sugahara ◽

Yuko Ogawa ◽

Suzue Furukawa ◽

Yuji Ishida ◽

...

Keyword(s):

Drug Development ◽

Drug Efficacy ◽

Developed Countries ◽

Liver Disorder ◽

New Drugs ◽

Peroxisome Proliferator ◽

Chimeric Mice ◽

Peroxisome Proliferator Activated Receptor ◽

Nonalcoholic Fatty Liver ◽

Nash Model

Nonalcoholic fatty liver disease/steatohepatitis (NAFLD/NASH) is the most common liver disorder in developed countries. Although many new therapeutics for NASH are present in the drug development pipeline, there are still no approved drugs. One of the reasons that makes NASH drug development challenging is the lack of appropriate animal NASH models that resolve issues arising from inter-species differences between humans and rodents. In the present study, we developed a choline-deficient, L-amino-acid-defined, high-fat-diet (CDAHFD)-induced human NASH model using human liver chimeric mice. We demonstrated human hepatocyte injury by an elevation of plasma human alanine aminotransferase 1 in mice fed CDAHFD. Histological analysis showed that CDAHFD feeding induced similar histological changes to human NASH patients, including ballooning, inflammation, apoptosis, regeneration of human hepatocytes, and pericellular and perisinusoidal fibrosis. The chimeric mice fed CDAHFD were treated with a peroxisome-proliferator-activated receptor α/δ agonist, Elafibranor. Elafibranor ameliorated steatosis, ballooning of hepatocytes, and preserved fibrosis progression. We developed a novel humanized NASH model that can elucidate pathophysiological mechanisms and predict therapeutic efficacy in human NASH. This model will be useful in exploring new drugs and biomarkers in the early stages of human NASH.

Download Full-text

MeFSAT: A curated natural product database specific to secondary metabolites of medicinal fungi

10.1101/2020.12.04.412502 ◽

2020 ◽

Author(s):

R.P. Vivek-Ananth ◽

Ajaya Kumar Sahoo ◽

Kavyaa Kumaravel ◽

Karthikeyan Mohanraj ◽

Areejit Samal

Keyword(s):

Secondary Metabolites ◽

Physicochemical Properties ◽

Natural Product ◽

Chemical Space ◽

Chemical Structures ◽

Therapeutic Uses ◽

Natural Product Library ◽

Medicinal Fungi ◽

Similarity Networks ◽

Shape Complexity

AbstractFungi are a rich source of secondary metabolites which constitutes a valuable and diverse chemical space of natural products. Medicinal fungi have been used in traditional medicine to treat human ailments for centuries. To date, there is no devoted resource on secondary metabolites and therapeutic uses of medicinal fungi. Such a dedicated resource compiling dispersed information on medicinal fungi across published literature will facilitate ongoing efforts towards natural product based drug discovery. Here, we present the first comprehensive manually curated database on Medicinal Fungi Secondary metabolites And Therapeutics (MeFSAT) that compiles information on 184 medicinal fungi, 1830 secondary metabolites and 149 therapeutics uses. Importantly, MeFSAT contains a non-redundant in silico natural product library of 1830 secondary metabolites along with information on their chemical structures, computed physicochemical properties, drug-likeness properties, predicted ADMET properties, molecular descriptors and predicted human target proteins. By comparing the physicochemical properties of secondary metabolites in MeFSAT with other small molecules collections, we find that fungal secondary metabolites have high stereochemical complexity and shape complexity similar to other natural product libraries. Based on multiple scoring schemes, we have filtered a subset of 228 drug-like secondary metabolites in MeFSAT database. By constructing and analyzing chemical similarity networks, we show that the chemical space of secondary metabolites in MeFSAT is highly diverse. The compiled information in MeFSAT database is openly accessible at: https://cb.imsc.res.in/mefsat/.

Download Full-text

Ship Classification in High-Resolution SAR Images Using Deep Learning of Small Datasets

Sensors ◽

10.3390/s18092929 ◽

2018 ◽

Vol 18 (9) ◽

pp. 2929 ◽

Cited By ~ 14

Author(s):

Yuanyuan Wang ◽

Chao Wang ◽

Hong Zhang

Keyword(s):

Deep Learning ◽

High Resolution ◽

Classification Accuracy ◽

Experimental Results ◽

Fine Tuning ◽

Classification Model ◽

Great Success ◽

Sar Images ◽

Convolutional Networks ◽

Ship Classification

With the capability to automatically learn discriminative features, deep learning has experienced great success in natural images but has rarely been explored for ship classification in high-resolution SAR images due to the training bottleneck caused by the small datasets. In this paper, convolutional neural networks (CNNs) are applied to ship classification by using SAR images with the small datasets. First, ship chips are constructed from high-resolution SAR images and split into training and validation datasets. Second, a ship classification model is constructed based on very deep convolutional networks (VGG). Then, VGG is pretrained via ImageNet, and fine tuning is utilized to train our model. Six scenes of COSMO-SkyMed images are used to evaluate our proposed model with regard to the classification accuracy. The experimental results reveal that (1) our proposed ship classification model trained by fine tuning achieves more than 95% average classification accuracy, even with 5-cross validation; (2) compared with other models, the ship classification model based on VGG16 achieves at least 2% higher accuracies for classification. These experimental results reveal the effectiveness of our proposed method.

Download Full-text

Pharmacogenomics Genome Wise Association Clinical Studies

Software Innovations in Clinical Drug Development and Safety - Advances in Medical Technologies and Clinical Practice ◽

10.4018/978-1-4666-8726-4.ch010 ◽

2015 ◽

pp. 218-233

Author(s):

Udayaraja GK

Keyword(s):

Side Effects ◽

Drug Toxicity ◽

Dose Adjustment ◽

Drug Efficacy ◽

New Drugs ◽

Genomic Research ◽

Drug Doses ◽

Genetic Makeup ◽

Trial And Error Method ◽

Error Method

Pharmacogenomics deals with drug responses in individual based on genetic variation in genome. Based on genetic variations, drugs may produce more or less therapeutic effect, and same way in side effects also. Physicians can use information about your genetic makeup to choose those drugs and drug doses to get better therapy. Optimizing drug therapy and rational dose adjustment with respect to genetic makeup will maximize drug efficacy and minimal adverse effects. This broken traditional ‘trial and error' method of ‘one drug fits all', and ‘one dose fits all' which contributing to 25–50% of drug toxicity or treatment failures. This will contribute to improve the ways in which existing drugs are used, genomic research will lead to drug development to produce new drugs that are highly effective without serious side effects. This approach to bring personalized medicine more practice and drug combinations are optimized for each individual' genetic makeup.

Download Full-text

Problems of the Fight Against Doping in Professional Sports

Advances in Business Information Systems and Analytics - Law, Ethics, and Integrity in the Sports Industry ◽

10.4018/978-1-5225-5387-8.ch010 ◽

2019 ◽

pp. 202-240

Author(s):

Anatoly Peskov

Keyword(s):

Sports Medicine ◽

New Technologies ◽

Professional Sports ◽

Administrative Law ◽

Therapeutic Use ◽

New Drugs ◽

Professional Athletes ◽

International Standard ◽

Main Factors ◽

Enforcement Mechanisms

Doping became, as many experts note, not only more diverse, highly specialized, and efficient, but also dangerous for the health of athletes. One of the main factors that allows athletes to escape responsibility is corruption. The author pays particular attention to research and new technologies in the field of sports medicine, including generating new kinds of doping. The chapter also examines the practice of international standard on granting exceptions on therapeutic use of drugs. The author suggests reconstructing the existing system of criminal and administrative law to develop new enforcement mechanisms in the fight against doping to impose a ban on the testing of new drugs on professional athletes.

Download Full-text

Adaptive CU Split Decision Based on Deep Learning and Multifeature Fusion for H.266/VVC

Scientific Programming ◽

10.1155/2020/8883214 ◽

2020 ◽

Vol 2020 ◽

pp. 1-11

Author(s):

Jinchao Zhao ◽

Yihan Wang ◽

Qiuwen Zhang

Keyword(s):

Deep Learning ◽

Video Coding ◽

High Efficiency ◽

Texture Classification ◽

Rate Distortion ◽

Classification Model ◽

High Efficiency Video Coding ◽

Fast Encoding ◽

Training Samples ◽

Coding Unit

With the development of technology, the hardware requirement and expectations of user for visual enjoyment are getting higher and higher. The multitype tree (MTT) architecture is proposed by the Joint Video Experts Team (JVET). Therefore, it is necessary to determine not only coding unit (CU) depth but also its split mode in the H.266/Versatile Video Coding (H.266/VVC). Although H.266/VVC achieves significant coding performance on the basis of H.265/High Efficiency Video Coding (H.265/HEVC), it causes significantly coding complexity and increases coding time, where the most time-consuming part is traversal calculation rate-distortion (RD) of CU. To solve these problems, this paper proposes an adaptive CU split decision method based on deep learning and multifeature fusion. Firstly, we develop a texture classification model based on threshold to recognize complex and homogeneous CU. Secondly, if the complex CUs belong to edge CU, a Convolutional Neural Network (CNN) structure based on multifeature fusion is utilized to classify CU. Otherwise, an adaptive CNN structure is used to classify CUs. Finally, the division of CU is determined by the trained network and the parameters of CU. When the complex CUs are split, the above two CNN schemes can successfully process the training samples and terminate the rate-distortion optimization (RDO) calculation for some CUs. The experimental results indicate that the proposed method reduces the computational complexity and saves 39.39% encoding time, thereby achieving fast encoding in H.266/VVC.

Download Full-text

A Novel Autonomous Perceptron Model for Pattern Classification Applications

Entropy ◽

10.3390/e21080763 ◽

2019 ◽

Vol 21 (8) ◽

pp. 763 ◽

Cited By ~ 12

Author(s):

Alaa Sagheer ◽

Mohammed Zidan ◽

Mohammed M. Abdelsamea

Keyword(s):

Pattern Classification ◽

Data Science ◽

Science Research ◽

Efficient Solutions ◽

Classification Model ◽

Computational Time ◽

Quantum Bit ◽

Training Samples ◽

Artificial Neural Network Ann ◽

Definition Of

Pattern classification represents a challenging problem in machine learning and data science research domains, especially when there is a limited availability of training samples. In recent years, artificial neural network (ANN) algorithms have demonstrated astonishing performance when compared to traditional generative and discriminative classification algorithms. However, due to the complexity of classical ANN architectures, ANNs are sometimes incapable of providing efficient solutions when addressing complex distribution problems. Motivated by the mathematical definition of a quantum bit (qubit), we propose a novel autonomous perceptron model (APM) that can solve the problem of the architecture complexity of traditional ANNs. APM is a nonlinear classification model that has a simple and fixed architecture inspired by the computational superposition power of the qubit. The proposed perceptron is able to construct the activation operators autonomously after a limited number of iterations. Several experiments using various datasets are conducted, where all the empirical results show the superiority of the proposed model as a classifier in terms of accuracy and computational time when it is compared with baseline classification models.

Download Full-text

Flavonoids from Brazilian Cerrado: Biosynthesis, Chemical and Biological Profile

Molecules ◽

10.3390/molecules24162891 ◽

2019 ◽

Vol 24 (16) ◽

pp. 2891 ◽

Cited By ~ 1

Author(s):

Josana de Castro Peixoto ◽

Bruno Junior Neves ◽

Flávia Gonçalves Vasconcelos ◽

Hamilton Barbosa Napolitano ◽

Maria Gonçalves da Silva Barbalho ◽

...

Keyword(s):

Plant Species ◽

Biological Activities ◽

Relevant Literature ◽

Phenylpropanoid Pathway ◽

New Drugs ◽

Brazilian Cerrado ◽

Biological Profile ◽

Main Route ◽

Chemical Structures ◽

Low Toxicity

Flavonoids are highly bioactive compounds with very low toxicity, which makes them attractive starting points in drug discovery. This study aims to provide information on plant species containing flavonoids, which are found in the Brazilian Cerrado. First, we present the characterization and plant diversity with emphasis on the families of flavonoid-producing plants, and then we describe the phenylpropanoid pathway which represents the flavonoids’ main route biosynthesis—generally conserved in all species. Chemical structures and biological activities of flavonoids isolated from the Cerrado’s plant species are also described based on examples from the relevant literature studies. Finally, research on the biodiversity of the Cerrado biome should be encouraged, due to the discovery of new sources of flavonoids which can provide several benefits to human health and the possibility of developing new drugs by the pharmaceutical industry.

Download Full-text