I analyzed two SNPs for linkage disequilibrium using SNPStats. It gave D'=0.9995 and r=0.99 but when I analyze my SNP data Manually, always these two SNPs are either present or absent together in all of my study subjects, giving a hint that these two SNPs are completely linked. Can I call these two SNPs in 'perfect linkage disequilibrium' despite D' and r values being very slightly less than 1. Do we ever get D' and r values equal to 1 practically? Do these values become closer to 1 for SNPs in perfect linkage disequilibrium with increasing sample size?
Of course, the higher the sample size, the higher the confidence into D' and r values, by reducing the probability to miss a recombination event. However, there is no reason that sample size influences D' and r values directly, since they are computed from exact frequencies from your sample. What you got is is a complete and perfect DL and the error to 1 is probably a problem of rounding intermediate results in the software. You can check this by directly computing D' and r from formulas.
- Badges
- Science method