more migration direction confusion

203 views
Skip to first unread message

Peter

unread,
May 18, 2012, 10:47:33 AM5/18/12
to Isolation with Migration
This IMa2 output header at the top of the 2Nm posterior estimates has
me confused because it seems inconsistent with the manual:

HISTOGRAM GROUP 4: POPULATION MIGRATION (2NM) POSTERIOR PROBABILITY
HISTOGRAMS
-----------------------------------------------------------------------------
curve height is an estimate of the posterior probability
each term is the product of a population parameter (e.g. q0) and a
migration rate (e.g.m0>1)
migration rates are in the coalescent (backwards in times), so that a
population migration rate of
q1m0>1 is the population rate (forward in time) at which population 1
receives migrants from population 0

IMa2 Manual p. 12: For example, m0>1 is the name of the migration rate
parameter for migration of genes from population 0 to 1, backwards in
time, in the coalescent direction. Interpreted forward in time, in the
usual way that most people think about migration, m0>1 is the rate at
which population 0 receives genes from population 1.

I'm stuck on the last line of the header from the output because if
q0m0>1 is the population migration rate for migrants going from
population 0 to population 1 backwards in time, I feel like that
header should read:

q1m1>0 is the population rate (forward in time) at which population 1
receives migrants from population 0.

jhey

unread,
Jun 12, 2012, 12:37:38 AM6/12/12
to Isolation with Migration
I think it is correct. The population migration rates (i.e. a
product of a population size parameter and a migration parameter)
refer to migration forward in time.

jhey

Peter

unread,
Jun 29, 2012, 10:10:36 AM6/29/12
to isolation-wi...@googlegroups.com
Ok, so you're saying the notation in the header of the output (the direction of the arrows in q1m0>1 ) is in the context of forward time because the 2Nm parameter only exists in the context of forward time?   

The manual has always made perfect sense but I think this part of the output is confusing because it's the only place where you give migration arrows in forward time; plus in the manual you've got this:

"Interpreted forward in time, in the usual way that most people think about migration, m0>1 is the rate at which population 0 receives genes from population 1." 

Anyway, not trying to be critical, but for the typical user I think it might be easier to keep migration straight if the arrows in the documentation and output were always in backward time?  
Reply all
Reply to author
Forward
0 new messages