szhan / tsimpute

Genome-wide genotype imputation using tree sequences.
MIT License
0 stars 0 forks source link

Update after getting imputation results using tsinferred trees #170

Closed szhan closed 10 months ago

szhan commented 10 months ago

Including the results from this run, we have learned the following:

Mean wrongly imputed sites

Imputation is best when doing it a la BEAGLE and worst when getting Viterbi paths on tsinferred trees.

Note that in these runs, we are using the same genome-wide recombination and mutation rates.