How to determine the subclade contributing to reproductive number increasing

[Jie Zha]

Dear Experts,

 I have estimated the trajectory distributions of reproductive numbers of three genotypes of viruese, using BDSKY model, and sumarized the corresponding MAP trees from the  the posterior tree samples generated from BDSKY model.  I identified the reproductive number increasing above 1 by checking the plots of the reproductive numbers,  for all the three genotypes, and then I want to identified the subclade of the MAP trees that contribute to the  increasings of the trajectory distributions of the reproductive numbers.  To achieve this goal,  I follow the principle as below: 

    The MAP trees were aligned to the time axis of the reproductive number skyline plots by node ages of the tree. When the interval of 95% HPD all above 1, the reproductive numbers were considered being significantly larger than 1, and the start times of the increasing phases before reaching the maximum reproductive numbers, were determined as the times when the median value of the reproductive numbers happen to exceed above 1. The subclades of the MAP tree will be considered contributing to reproductive number becoming significantly larger than 1, if their node ages are equal to or a bit larger than the start time values of the increasing phases.

I attached the three plots of the three genotypes viruses below:

In the first two plots we can see the start time values are amost equal to the node ages of the subclades that are probably contributing to the reproductive number increasings, but in the third plot, we can see the start time is significantly smaller than the the node ages of the subclades.

I am not sure whether  the principle I used to identify the subclades, is appropriate or not?  If it is appropriate, why the start time is significantly smaller than the the node ages of the subclades for the third genotypes of virses.

Best,

Jie