Limits...
A network-based classification model for deriving novel drug-disease associations and assessing their molecular actions.

Oh M, Ahn J, Yoon Y - PLoS ONE (2014)

Bottom Line: The growing number and variety of genetic network datasets increases the feasibility of understanding how drugs and diseases are associated at the molecular level.These quantified scores were used as features for the prediction of novel drug-disease associations.Our classifiers using Random Forest, Multilayer Perceptron and C4.5 showed a high specificity and sensitivity (AUC score of 0.855, 0.828 and 0.797 respectively) in predicting novel drug indications, and displayed a better performance than other methods with limited drug and disease properties.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Engineering, Gachon University, Seongnam, Korea.

ABSTRACT
The growing number and variety of genetic network datasets increases the feasibility of understanding how drugs and diseases are associated at the molecular level. Properly selected features of the network representations of existing drug-disease associations can be used to infer novel indications of existing drugs. To find new drug-disease associations, we generated an integrative genetic network using combinations of interactions, including protein-protein interactions and gene regulatory network datasets. Within this network, network adjacencies of drug-drug and disease-disease were quantified using a scored path between target sets of them. Furthermore, the common topological module of drugs or diseases was extracted, and thereby the distance between topological drug-module and disease (or disease-module and drug) was quantified. These quantified scores were used as features for the prediction of novel drug-disease associations. Our classifiers using Random Forest, Multilayer Perceptron and C4.5 showed a high specificity and sensitivity (AUC score of 0.855, 0.828 and 0.797 respectively) in predicting novel drug indications, and displayed a better performance than other methods with limited drug and disease properties. Our predictions and current clinical trials overlap significantly across the different phases of drug development. We also identified and visualized the topological modules of predicted drug indications for certain types of cancers, and for Alzheimer's disease. Within the network, those modules show potential pathways that illustrate the mechanisms of new drug indications, including propranolol as a potential anticancer agent and telmisartan as treatment for Alzheimer's disease.

Show MeSH

Related in: MedlinePlus

System overview.(a) “Adjacency-Based Inference” measures the drug-drug (disease-disease) adjacency among known drug-disease associations, and infers new drug-disease association. “Module-Distance-Based Inference” derives drug-drug (disease-disease) gene module among known drug-disease associations, measures the distance between the gene module and disease (drug), and infers new drug-disease association. (b) Drug-disease relationship represented by score becomes features. Various machine learning based classifiers are built with those features, and predict unknown drug-disease relationship.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4214731&req=5

pone-0111668-g001: System overview.(a) “Adjacency-Based Inference” measures the drug-drug (disease-disease) adjacency among known drug-disease associations, and infers new drug-disease association. “Module-Distance-Based Inference” derives drug-drug (disease-disease) gene module among known drug-disease associations, measures the distance between the gene module and disease (drug), and infers new drug-disease association. (b) Drug-disease relationship represented by score becomes features. Various machine learning based classifiers are built with those features, and predict unknown drug-disease relationship.

Mentions: The proposed method consists of two processes, as shown in Figure 1. In the first stage, the degree of the drug-disease association is scored by means of adjacency-based inference and module-distance-based inference. Detailed descriptions of the adjacency-based inference and module-distance-based inference methods are given in section “Two methods for scoring drug-disease associations”. In the second stage, the scores from the first stage are regarded as features characterizing the drug-disease relationship; a classifier is subsequently built using these features by means of learning. With this classifier, predictions are made regarding whether an unknown drug-disease pair has an association. Finally, new drug-disease associations are discovered. The details of this stage are given in section “Characterizing a drug-disease relationship via features”.


A network-based classification model for deriving novel drug-disease associations and assessing their molecular actions.

Oh M, Ahn J, Yoon Y - PLoS ONE (2014)

System overview.(a) “Adjacency-Based Inference” measures the drug-drug (disease-disease) adjacency among known drug-disease associations, and infers new drug-disease association. “Module-Distance-Based Inference” derives drug-drug (disease-disease) gene module among known drug-disease associations, measures the distance between the gene module and disease (drug), and infers new drug-disease association. (b) Drug-disease relationship represented by score becomes features. Various machine learning based classifiers are built with those features, and predict unknown drug-disease relationship.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4214731&req=5

pone-0111668-g001: System overview.(a) “Adjacency-Based Inference” measures the drug-drug (disease-disease) adjacency among known drug-disease associations, and infers new drug-disease association. “Module-Distance-Based Inference” derives drug-drug (disease-disease) gene module among known drug-disease associations, measures the distance between the gene module and disease (drug), and infers new drug-disease association. (b) Drug-disease relationship represented by score becomes features. Various machine learning based classifiers are built with those features, and predict unknown drug-disease relationship.
Mentions: The proposed method consists of two processes, as shown in Figure 1. In the first stage, the degree of the drug-disease association is scored by means of adjacency-based inference and module-distance-based inference. Detailed descriptions of the adjacency-based inference and module-distance-based inference methods are given in section “Two methods for scoring drug-disease associations”. In the second stage, the scores from the first stage are regarded as features characterizing the drug-disease relationship; a classifier is subsequently built using these features by means of learning. With this classifier, predictions are made regarding whether an unknown drug-disease pair has an association. Finally, new drug-disease associations are discovered. The details of this stage are given in section “Characterizing a drug-disease relationship via features”.

Bottom Line: The growing number and variety of genetic network datasets increases the feasibility of understanding how drugs and diseases are associated at the molecular level.These quantified scores were used as features for the prediction of novel drug-disease associations.Our classifiers using Random Forest, Multilayer Perceptron and C4.5 showed a high specificity and sensitivity (AUC score of 0.855, 0.828 and 0.797 respectively) in predicting novel drug indications, and displayed a better performance than other methods with limited drug and disease properties.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Engineering, Gachon University, Seongnam, Korea.

ABSTRACT
The growing number and variety of genetic network datasets increases the feasibility of understanding how drugs and diseases are associated at the molecular level. Properly selected features of the network representations of existing drug-disease associations can be used to infer novel indications of existing drugs. To find new drug-disease associations, we generated an integrative genetic network using combinations of interactions, including protein-protein interactions and gene regulatory network datasets. Within this network, network adjacencies of drug-drug and disease-disease were quantified using a scored path between target sets of them. Furthermore, the common topological module of drugs or diseases was extracted, and thereby the distance between topological drug-module and disease (or disease-module and drug) was quantified. These quantified scores were used as features for the prediction of novel drug-disease associations. Our classifiers using Random Forest, Multilayer Perceptron and C4.5 showed a high specificity and sensitivity (AUC score of 0.855, 0.828 and 0.797 respectively) in predicting novel drug indications, and displayed a better performance than other methods with limited drug and disease properties. Our predictions and current clinical trials overlap significantly across the different phases of drug development. We also identified and visualized the topological modules of predicted drug indications for certain types of cancers, and for Alzheimer's disease. Within the network, those modules show potential pathways that illustrate the mechanisms of new drug indications, including propranolol as a potential anticancer agent and telmisartan as treatment for Alzheimer's disease.

Show MeSH
Related in: MedlinePlus