The impact of genome-wide association studies on biomedical research publications

Abstract

The impact of genome-wide association studies (GWAS) on biomedical research output was quantitatively evaluated. GWAS have not changed the historical skew of publications toward genes involved in Mendelian diseases, but genes newly implicated by GWAS in complex disease do experience a modest increase in publication activity. The impact of GWAS on biomedical research into individual genes is declining, however, even as the rate of new GWAS associations increases.A decade after the first publication of a successful GWAS [1], few results from GWAS have had clinical impact, because most associated variants have modest effect sizes or unclear functional consequences [2,3]. Direct clinical impact is, however, not the only goal of GWAS. Another major goal is to identify and steer research toward novel genes involved in complex diseases [4]. For example, the first published GWAS unexpectedly identified Complement Factor H as associated with macular degeneration [1], spurring the development of complementbased therapeutics [5]. Similarly, most genes associated with multiple sclerosis through GWAS had not previously been considered candidates [6]. But what impact have GWAS had beyond these paradigm examples? GWAS are more highly cited than comparable candidate gene studies [7], but how much follow-up research do GWAS actually motivate? To answer this question, we quantified the effects of GWAS on biomedical research output.We measured research output on genes using scientific publications, as collected in the NCBI Gene database [8]. We prefer this manually curated database to automatic text mining, because text mining may introduce false positives when a gene is mentioned in passing. We classified genes into those associated with Mendelian disease, complex disease, both, or no disease using the Online Mendelian Inheritance in Man (OMIM) database [9] and the EBI-NCBI GWAS catalog [10].As expected [11,12], we found that in the pre-GWAS era research was skewed toward a minority of human genes (Fig. 1A). The majority of highly-studied genes were involved in Mendelian disease, and many genes that would later be associated with complex disease received little attention (Fig. 1B). In the post-GWAS era, research output is even more skewed ( Fig. 1C; coefficient of variation 3.4 vs 2.5). Most highly studied genes are still ones involved in Mendelian disease, and many genes associated with complex disease still receive little attention (Fig. 1D). By contrast, research output on yeast genes became much less skewed following the publication of the yeast genome (Fig. S1). The advent of GWAS has not reduced the bias of biomedical research toward Mendelian disease genes. But how has GWAS affected research on individual genes?To quantify the immediate effect of GWAS on individual "hit" genes, we focused on the year a gene was associated with complex disease through GWAS and the A: In the pre-GWAS era (before 2005), the distribution of publications among genes was highly skewed. B: Highlystudied genes tended to be involved in Mendelian di...