Hello all,
I have population genetics data (mtDNA) for 5 populations (n=20) of the same species. All neutrality tests (Fu's Fs, Tajima's D, Fu & Li's F, Fu & Li's D, Roza's R2, Strobeck's S) are non-significant so no signal of deviation from neutrality is there. However mismatch distribution analysis (with demographic expansion model) for 3 of the 2 populations are unimodal, the other two are multimodal. SSD for multimodal distributions are significant and the rest are not, suggesting three of the populations are not significantly different than population expansion model. Raggedness index for all of the populations are non-significant suggesting all of the populations are not significantly different than population expansion model.
1) What could cause the contrasting results from neutrality tests and mismatch analysis?
2) Why is raggedness index failing to capture the multi-modal distribution of the two populations while SSD can?
Another contradiction I face is between the actual mismatch distribution and the mismatch statistics. I attached a mismatch distribution for a different population. It is unimodal (albeti skewed) and observed values seem like a real good match to the simulated values. However SSD (0.00236) and raggedness index (0.07244) for this plot are both significant.
I am a bit lost in what to deduce from these results. Can you help me please?
Thanks!
Do I understand the question correctly that all neutrality tests except raggedness are not deviating from their respective null hypothesis, i.e. all five populations are inferred to be at equilibrium according to those tests?
For raggedness, 3 out of 5 are unimodal, but not significantly so; and 2 are significantly multimodal?
Multimodality points in the same direction as the other estimators, populations at equilibrium.
So the problem would be the ones with unimodal distributions of pairwise differences among haplotypes, with a smooth distribution, which would indicate recent population expansion in contrast to all other estimators.
According to Ramos-Onsins and Rozas (2002), the most powerful tests for detecting population growth are Fu’s FS test and their R2 test, whereas the raggedness is relatively weak. (Ramos-Onsins and Rozas, 2002. Statistical Properties of New Neutrality Tests Against Population Growth. Mol. Biol. Evol.19:2092–2100)
So if the unimodality is not significant, and the estimator is weak, you perhaps wouldn't have to put that much trust in those particular results?
Can anybody help me how can I interpret my opposite results between neutrality test-mismatch curve and SSD-raggedness index (Hri) tests?
In my results, insignificant Tajima’s D= -0.76 (P= 0.24); Fu’s Fs= -1.59 (P= 0.27) and multi modal observed MMD curve denotes populations are at equilibrium. However, SSD = 0.016 (P = 0.48) and Hri = 0.009 (P = 0.15) are insignificant which denotes population expansion.
Please see the attached picture. I used Arlequin program for analysis. In my d-loop DNA sequence, I have long indel zones in the sequences.
Can anybody help me, please?
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology