Feature request: allow multiple MRCA priors

[richelbilderbeek]

From another Issue, a suggestion by @evolucionario:

Implementing the option of calibrating multiple nodes would also be great. I would immediately use it in the simulations I'm performing.

[richelbilderbeek]

I think adding multiple MRCA priors:

• Allows for a better idea of adding multiple tree/clock/site priors
• Is rather easy
• Does break the build, as the mrca_prior argument will be renamed to mrca_priors, but it will not result in users having to rewite their code except for adding the s, as suggested by a stop/warning
• I do need a volunteer to convince me to do so

[richelbilderbeek]

Sent an email to @evolucionario: maybe a student can add it.

[richelbilderbeek]

Sent an email to @joelnitta: maybe a student can add it.

[joelnitta]

Sorry, I'm not aware of anybody (myself included) who can work on this at the moment.

[mmatschiner]

Thanks for developing babette, a tool like this has long been missing, and I'm already using babette a lot. But I just realized that it doesn't seem to be possible to use more than one mrca prior. If that is so, it would be great if the option could be added. It also seems that the option used to be available, because create_beast2_input_file used to have the option mrca_priors which is now deprecated? Funny that I'm running into this issue just a day after you've opened it here. If this could be added within the next week or so, I would implement it in my current pipeline; if not, I'll move on with an alternative solution.

[mmatschiner]

Also funny that the issue has been raised by @evolucionario as I wanted to use this to implement constraints from a paper by him...

[richelbilderbeek]

Sent an email to @mmatschiner: maybe a student can add it.

[richelbilderbeek]

I may have found someone willing to do this.

Plan:

• Create some tests that should work together, including the test that literally reproduces BEAUti output
• Demonstrate fixing tests