Benchmarking network-based gene prioritization methods for cerebral small vessel disease.

Huayu Zhang,Keith Smith,Muchen Jiang,Amy Ferguson,Cathie Sudlow,Teng Zhang,Grant Robertson,Kristiina Rannikmae,Honghan Wu

doi:10.1093/bib/bbab006

Abstract

Network-based gene prioritization algorithms are designed to prioritize disease-associated genes based on known ones using biological networks of protein interactions, gene–disease associations (GDAs) and other relationships between biological entities. Various algorithms have been developed based on different mechanisms, but it is not obvious which algorithm is optimal for a specific disease. To address this issue, we benchmarked multiple algorithms for their application in cerebral small vessel disease (cSVD). We curated protein–gene interactions (PGIs) and GDAs from databases and assembled PGI networks and disease–gene heterogeneous networks. A screening of algorithms resulted in seven representative algorithms to be benchmarked. Performance of algorithms was assessed using both leave-one-out cross-validation (LOOCV) and external validation with MEGASTROKE genome-wide association study (GWAS). We found that random walk with restart on the heterogeneous network (RWRH) showed best LOOCV performance, with median LOOCV rediscovery rank of 185.5 (out of 19 463 genes). The GenePanda algorithm had most GWAS-confirmable genes in top 200 predictions, while RWRH had best ranks for small vessel stroke-associated genes confirmed in GWAS. In conclusion, RWRH has overall better performance for application in cSVD despite its susceptibility to bias caused by degree centrality. Choice of algorithms should be determined before applying to specific disease. Current pure network-based gene prioritization algorithms are unlikely to find novel disease-associated genes that are not associated with known ones. The tools for implementing and benchmarking algorithms have been made available and can be generalized for other diseases.

Highlights

‘Guilt by association’ is the most adopted concept in networkbased gene prioritization methods
We found that random walk with restart on the heterogeneous network (RWRH) showed best leave-one-out cross-validation (LOOCV) performance, with median LOOCV rediscovery rank of 185.5
The underlying principle is that genes that are closely associated in the protein–gene interaction (PGI) network tend to be in the same functional module, thereby giving rise to similar phenotypes [1]

Summary

Introduction

‘Guilt by association’ is the most adopted concept in networkbased gene prioritization methods. Different algorithms have been developed and applied to biological interaction networks under this principle These algorithms take a set of known genes associated with a disease (seed genes) as input and try to predict or prioritize other potential genes associated with the disease. The reports describing the algorithms typically showcased their performance in an example disease or condition, so that it is not clear for end users who wish to apply the algorithms to the disease of their interest which algorithm is the optimal one. To address this issue, we benchmarked seven representative algorithms for their application in nonamyloid cerebral small vessel disease (hereafter referred to as cSVD). The candidate genes prioritized by best performing algorithms were externally evaluated with results of genome-wide association study (GWAS) MEGASTROKE [15, 16]

Methods

Results

Conclusion