over estimating tau and Ne

[gargkritika]

Dear Dr. Gronau,

I am using G-PhoCS for a radiation of fruit bats. I was able to successfully run G-PhoCS. However, the Tau and Ne are being overestimate. I would expect divergence between species during the Pleistocene, however, the estimates I am getting are around 700 million years. Similarly Ne is also being overestimated by a 1000 fold. I am estimating divergence time by dividing the estimates of tau by the mutation rate per gen. I have run models both with and without gene flow and get the same issues.

Could you please suggest a possible solution. Do let me know if you need access to any files.

Thank you.

Best regards,
Kritika

[gphocs-dev]

Dear Kritika, Did you divide by the "print factor" set in the control file. All raw theta and tau estimates are scaled by a constant factor to avoid loss of accuracy when printing small values. This is typically set to 10,000, but you should check your control file.

…

--Ilan

On Mon, Aug 17, 2020 at 8:06 AM gargkritika ***@***.***> wrote: Dear Dr. Gronau, I am using G-PhoCS for a radiation of fruit bats. I was able to successfully run G-PhoCS. However, the Tau and Ne are being overestimate. I would expect divergence between species during the Pleistocene, however, the estimates I am getting are around 700 million years. Similarly Ne is also being overestimated by a 1000 fold. I am estimating divergence time by dividing the estimates of tau by the mutation rate per gen. I have run models both with and without gene flow and get the same issues. Could you please suggest a possible solution. Do let me know if you need access to any files. Thank you. Best regards, Kritika — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#70>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADO7ILUXTP7AX2NN54MNCFTSBC3G3ANCNFSM4QBHZB5A> .

[gargkritika]

Dear Dr. Gronau,

Thank you so much. Makes sense now. I think this parameter is set when we use the java applet version to generate the control file and hence I missed it.

Thank you once again.

Best regards,
Kritika