The problem of parsing RELAX results

[1221li]

Dear Developer,

Thank you very much for developing such a useful software, which has provided me with great help. But I have some doubts about how to interpret this result.

If I have nine species, and I use one of the lineages as the "TEST" for the RELAX analysis. I'm going to get the json result, which has the K value for each species and node, and if I understand all right, a "relaxation or intensification parameter", which seems to be the final K value. My question is, should I use the final K value or can I also use the k value for each species and node. Whether the results of labeling multiple species as TEST differ substantially from those of labeling a single species as TEST.

Looking forward to your reply!

[spond]

Dear @1221li,

In your actual example, the p-value for the test is not significant (p~0.4), hence K is not significantly different from 1.

RELAX will fit several models and each may have different K estimates. Which one you use depends on your application. The standard use of RELAX is to report the overall K value (test vs reference), but only if the p-value shows a significant result.

Best,
Sergei

PS. Consider loading the JSON into the RELAX visualization module at vision.hyphy.org/RELAX