Postprocessing - Plotmap Bugs

[GabrieleNocchi]

Hi,

I have found a few possible bugs in the function plotMap. I can't provide any sample of my data at the moment however I double checked using a very small subset and gds file to make sure it was not a problem with my original data files.

I found 2 possible bugs:

1. The maptype marker (absence and presence marker plot) does not always retrieve the correct genotype of the marker for homozygous calls (homozygous reference and homozygous alternative). It seems to be consistent in retrieving the heterozygous calls correctly. I tried to look at the code but I cannot debug.
   Example: if you have a SNP with ID SNP_ID, biallelic with 2 possible states for example (A and T). If you do the presence absence map for the marker SNP_ID_AT it does it correctly. But for SNP_ID_AA or SNP_ID_TT it does not (most of the times).
   The homozygous calls often get mixed up or both return the same plot (same plot for either AA or TT). Sometimes it works by just adding an unexistent genotype for the marker, for example if SNP_ID_AA and SNP_ID_TT display the same absence/presence plot (possibly for example for genotype AA). Sometimes I managed to display the other homozygous genotype (TT in this example) by just typing a completely different allele in markerName (for example SNP_ID_CC). There is something wrong in the way plot map retrieves the genotype from the GDS file (I checked my ped from where I create my GDS and it is correct) and it does not seem to check whether the genotype you inserted exists for that marker (as long as you use A T C or G as suffix in marer name ie. SNP_ID_GG, SNP_ID_AA, SNP_ID_CC, SNP_ID_TT all retrieve something).

2. The second bug I found in in the env distribution plot. Here again, I tried to debug but I cannot read your code well. I am not sure if the data is plotted using the data in the R-sambada-env input file or from the raster, but I believe it is from the file beacause the function seems to plot the env variable of your samples. Maximum and minimum values (judging by the intensity of the red) are not on the samples they are suppose to be and also it rescale every variable from -50 to 50 and it is hard to comprehend this plot.