millanek / Dsuite

Fast calculation of Patterson's D (ABBA-BABA) and the f4-ratio statistics across many populations/species
160 stars 26 forks source link

Question about order of population set when running Dtrios #83

Open hungweichen0327 opened 9 months ago

hungweichen0327 commented 9 months ago

Dear @millanek and community,

I have questions about the order of population set when running Dusite Dtrios. If I have 3 populations (P1, P2, P3) and an outgroup population, the two orders in the population set below would have a different meaning when running Dsuite Dtrios. Because the first one indicates (P1, P2), P3), O) and the second one indicates (P2, P3) P1, O). Is that correct?

(1) P1 P2 P3 O
(2) P2 P3 P1 O

When I ran the Dsuite Dtrios, there were only 56 output results (meaning 56 combinations) when I had 8 populations with an outgroup population. But the output results would have 336 (8x7x6).

Thus, should I use the -t option and give the tree file providing the relationship of 8 populations when running Dsuite Dtrios? The output results will contain (P2, P3) P1, O) rather than (P1, P2), P3), O) if the tree is (P1, P2), P3), O)?

Thank you for the help.

millanek commented 9 months ago

Yes, if the tree file specifies (P1, P2), P3), O) then that's what the _tree.txt output is going to contain. However, do double check that this is the "correct" tree -- if the _BBAA.txt output is giving you a different orientation, it is for a reason.

The meaning of the different output files "_tree.txt", "_BBAA.txt", and "_Dmin.txt" is explained the documentation here in GitHub and more in detail in the paper.

All the best Milan