Based on my previous post (#9510165), I'm wondering why there are certain genes that cannot be mapped when running GDCprepare on data downloaded via TCGAbiolinks? Everytime I run GDCprepare on my dataset, I get the following message:

From the 60483 genes we couldn't map 3881

I've been ignoring it because I assumed all the genes were simply outdated and no longer existed in the current version of ensembl. However, after looking deeper into it, I've found that some of these unmappable genes are actually very mappable.

There were 41 genes in the dataset that were not mapped, yet clearly exist, and I can map them easily enough myself. For example, ENSG00000198899 and ENSG00000228253. All of these 41 genes seem to be "Mitochondrially Encoded".

What I don't understand is why this is happening. Is it intentional? Is it known? Is there something weird with "Mitochondrially Encoded" genes that I'm unaware of?