bug in missingnessHeatMaps.R

[elbourret]

I found a bug in missingnessHeatMaps.R

On line 101, you need to replace the all.equal() function with the identical() function, because the former will not always return a boolean, and therefore if the two matrices are different, the correct error message will never be written.

https://github.com/atcg/clustOpt/blob/20033424acf3405f2d3e1af0a0045124426624df/resources/missingnessHeatMaps.R#L101C14-L101C23

[elbourret]

There is another issue in the same file.

On line 141, you need to set autosome.only=FALSE in the call to function snpgdsIBS(), otherwise it fails for assemblies that were assembled de novo with iPyrad for instance.

https://github.com/atcg/clustOpt/blob/20033424acf3405f2d3e1af0a0045124426624df/resources/missingnessHeatMaps.R#L141C32-L141C41

[elbourret]

The code also has trouble dealing with datasets where some sample pairs share no SNPs, which results in snpgdsIBS() to call NaN as their IBS. The problem can be resolved by inputting the missing (NaN) values with the average IBS. I have fork this script to show the changes I made to resolve this issue.