How choose the topology with more than 5 populations

60 views
Skip to first unread message

federica mattucci

unread,
Oct 2, 2013, 10:45:10 AM10/2/13
to pop...@googlegroups.com
Hi,
I'm using the software popABC aiming at inferring historical demographic parameters on 9 populations of European wildcats.
I have some difficulties on choosing the topology in the prs files because with 9 populations i cannot type one of the 4 options (1-choose topology;2-specify topology manually etc etc).
It appear only this possibility:

"choose the topology to be used (16 numbers) available populations (0 1 2 3 4 5 6 7 8) from...to "

so how can i do!

the topology i need for my populations is (((0,1,2), (3,4,5), (6,7)), 8) but i don't know how write it on!!!

Thank you very much for your availability!
with best reguards


Federica Mattucci

ISPRA - Istituto Superiore per la Protezione e la Ricerca Ambientale

40064 Ozzano Emilia Bologna (Bo), Italy


Joao Sollari Lopes

unread,
Oct 2, 2013, 11:53:21 AM10/2/13
to pop...@googlegroups.com
Hi Federica Mattucci,

Here is how you should do:

Your topology is

(((0,1,2), (3,4,5), (6,7)), 8)

you don't specify the order at which these populations join together (maybe that is not important), so I'll assume the following population-joining events

1->0, 2->0, 4->3, 5->3, 7->6, 3->0, 6->0, 8->0, where -> is "join with"

in the .prs file you should specify

[topology_type] [topology_scheme]

3 1 0 2 0 4 3 5 3 7 6 3 0 6 0 8 0, where the first number chooses the topology_type

if you want to compare different topologies you should mark your topologies in the .dat file by specifying it in the .prs file

[topology_type] [topology_scheme] [topology_marker]

5 1 0 2 0 4 3 5 3 7 6 3 0 6 0 8 0 99, where the first number chooses the topology_type and the last is the topology-marker

Then you have to join the .dat files together (using popabc, if you want) and perform the rejection-step on the combined .dat files.

Hope this helps,
Joao

federica mattucci

unread,
Oct 7, 2013, 12:26:26 AM10/7/13
to pop...@googlegroups.com
thank you very much!!!

Federica


Reply all
Reply to author
Forward
0 new messages