Risk of false positive genetic associations in complex traits with underlying population structure: a case study.
Authors: Finno Carrie J, Aleman Monica, Higgins Robert J, Madigan John E, Bannasch Danika L
Journal: Veterinary journal (London, England : 1997)
Summary
# Editorial Summary Genome-wide association (GWA) studies have successfully identified single-gene mutations in horses, but their application to complex inherited diseases remains problematic, particularly when sample sizes are limited and population stratification exists. Finno and colleagues investigated equine neuroaxonal dystrophy (NAD), a common neurological disorder, using 42,819 SNP markers in 99 Quarter horses (37 affected, 62 unaffected), then sequenced candidate genes within any identified regions and validated findings through quantitative RT-PCR and association testing. Although no genome-wide significant association was detected, a suggestive signal emerged on chromosome 8 when only the ten most severely affected horses were analysed; however, systematic sequencing of three candidate genes in this region (PIK3C3, RIT2, and SYT4) ruled them out as causative variants. This work highlights a critical limitation of current equine GWA methodology: studies using 40–50,000 markers with sample sizes of 30–40 individuals (particularly those with fewer than 15 affected animals) are at substantial risk of producing false positive associations, especially when underlying population structure is not controlled for. For equine professionals and researchers, the takeaway is that rigorous candidate gene validation and replication in larger, more genetically diverse populations are essential before accepting GWA findings, and that negative results like these are equally valuable for preventing the perpetuation of spurious associations in the literature.
Read the full abstract on PubMed
Practical Takeaways
- •Genetic associations reported from small case-control studies (particularly <50 affected animals) should be interpreted cautiously and require independent validation before clinical application
- •Population stratification in horse breeds can generate spurious associations that do not represent true disease loci; always verify candidate genes through functional studies before accepting findings
- •When evaluating NAD in Quarter horses, genetic testing based on unvalidated markers may provide false reassurance or incorrect prognostic information
Key Findings
- •A case-control GWA study of 99 Quarter horses (37 affected, 62 unaffected) using 42,819 SNPs failed to achieve genome-wide significance for NAD susceptibility loci
- •A suggestive association on chromosome 8 (positions 62130605-62134644, log(1/P)=5.56) emerged only when analysis was restricted to 10 most stringently phenotyped NAD-affected horses
- •Sequencing and association testing excluded variants in candidate genes PIK3C3, RIT2, and SYT4 as causative for equine neuroaxonal dystrophy
- •The study demonstrates that small sample sizes (10-37 affected individuals) with 40-50,000 SNP markers are vulnerable to false positive associations due to population structure in complex traits