Prediction and Validation of Gene-Disease Associations Using Methods Inspired by Social Network Analyses

Abstract

Correctly identifying associations of genes with diseases has long been a goal in biology. With the emergence of large-scale gene-phenotype association datasets in biology, we can leverage statistical and machine learning methods to help us achieve this goal. In this paper, we present two methods for predicting gene-disease associations based on functional gene associations and gene-phenotype associations in model organisms. The first method, the Katz measure, is motivated from its success in social network link prediction, and is very closely related to some of the recent methods proposed for gene-disease association inference. The second method, called Catapult (Combining dATa Across species using Positive-Unlabeled Learning Techniques), is a supervised machine learning method that uses a biased support vector machine where the features are derived from walks in a heterogeneous gene-trait network. We study the performance of the proposed methods and related state-of-the-art methods using two different evaluation strategies, on two distinct data sets, namely OMIM phenotypes and drug-target interactions. Finally, by measuring the performance of the methods using two different evaluation strategies, we show that even though both methods perform very well, the Katz measure is better at identifying associations between traits and poorly studied genes, whereas Catapult is better suited to correctly identifying gene-trait associations overall.The authors want to thank Jon Laurent and Kris McGary for some of the data used, and Li and Patra for making their code available. Most of Ambuj Tewari's contribution to this work happened while he was a postdoctoral fellow at the University of Texas at Austin.

Mining Multi Drug-Pathways via A Probabilistic Heterogeneous Network Multi-label Classifier

“…In addition, Yu et al [27] reported a systematic approach that efficiently integrates the chemical, genomic, and pharmacological information for multi drug targeting and discovery on a large scale, based on two powerful methods of Random Forest (RF) and Support Vector Machine (SVM). Finally, Singh-Blom et al [28] presented two methods for predicting gene-disease associations based on functional gene associations and gene-phenotype associations in model organisms, which is close to this work. Those methods are: the Katz measure, used in social network link prediction, and CATAPULT (Combining dATa Across species using Positive-Unlabeled Learning Techniques, a supervised machine learning method that uses a biased support vector machine, where the features are derived from walks in a heterogeneous gene-trait network.…”

Computational Drug Discovery with Dyadic Positive-Unlabeled Learning

“…There has been an increasing amount of works on recovering gene-disease associations using networkbased algorithms [34,17,27] Gonen and Kaski [10] proposed Kernelized Bayesian Matrix Factorization (KBMF) to predict drug-target interactions by making use of the information from multiple domains via kernel methods. The Multiple Similarities Collaborative Matrix Factorization (MSCMF) [36], was proposed for drug-target interaction prediction which approximates the indicator matrix by the product of projection matrices of drug and target similarity matrices.…”

Large-scale integration of heterogeneous pharmacogenomic data for identifying drug mechanism of action

“…For each test drug, we ran a weighted majority voting among its k = 10 most similar drugs based on the cosine distances imposed over the vector representations of drugs. Following the previous work, 26 we used the "recall@top-R" as the evaluation metric, which is defined as the fraction of true associated MoAs (or targets) that were retrieved in the list of top-R predictions for a drug. The motivation of using this metric was that a method that can recover the true MoAs (or targets) in the top-R predictions with high probability is desirable and useful in applications such as drug repurposing.…”