csmiguel / usat_snp Goto Github PK

Add to bottom half a similar plot for the Hyla data, though for H. molleri the SSC will have to be calculated on the average of the localities.
Hand draw lines englobing the clusters.

Silvia suggested to plot diversity on a map.

check clumpp does the job
build assertion for rownames from qclumpak
rownames from results of Pelobates are different from structure respect to gen because I changed them-> solve that
run clumpp with dapc results as well

200 SNPs have roughly the same info that the microsatellite dataset. In order to compare the behaviour of the curves at the plot dnp/dnr Craig suggested to subsample 200 snps instead of using the whole dataset.

• another option might be to predict in the model at which no. of snps the mean bs support for the snps intersects the average bs for usats.

1. Change the order of the panels: left snps and right usats.
2. Change color palette for the usats since they do not match the clusters from the snps. If using the same palette one tries to intuitively do the match, even when there is no match.

The metrics Dnp and Dnr are done over standarized tree distances. However, in the plots "DnrDnp_ggplot_bw.pdf" it is possible to see that metric for both markers do not have the same highest value. This could be caused by the fact that the lengths of the branches supporting the trees are different between usats and SNPs. For that reason, in the meeting from Sept 13th 2019, it was suggested to normalize the the distances using the maximum distance from any internal node (non-tip nodes) instead of normalizing as I have it now:
maxd <- ape::cophenetic.phylo(tr1) %>% max()
tr1$edge.length <- tr1$edge.length / maxd

Craig suggested to change the order of the sections in the manuscript. As it is now:

1. population structure, 2. phylogeny, 3. dnp/dnr, 4. diversity

Craig suggested:

1. phylogeny, 2. genetic diversity, 3. pop. structure 4. interaction dnp/dnr