And named it after the current president of Brazil.

Clearly much better than mapping against a current extant (descendant) genome where biases are correlated with the amount of divergence of the mapped and reference genomes.

I like the second idea. This way all the descendants are equally divergent from their ancestor and this any bias is going to be equally and randomly distributed in the descendants.

So the idea would be to choose based on phylogeny rather than on a statistic of performance (do the mapping against sensible choices [phylogenetically informed] and compare the metrics). What are the phylogenetic constrains of these mapping and population genetic metrics?

Just last week I saw this three legged Crocodilurus amazonicus in the central Amazon.

Wouldn’t expect anything else from their ilk.

NKDV Order No. 00447, July 30, 1937, Stalin’s USSR, the Great Purge