Hello all,
I am currently working on a quantitative genetic project looking at the
heritability of morphological traits in Atlantic cod. My aim is to estimate
heritability using the animal model in a Bayesian context. I have built the
animal model using WinBUGS, but when I try to run it, the model doesn’t seem to
ever want to stop compiling (it ran for more than 12 hrs and was still
compiling). In my model, some of the terms, for example the breeding value,
follow a multivariate normal (MVN) distribution with mean 0 and variance
A(sigma2A). Where A is the additive relationship matrix between each fish (0
for no relationship, 0.25 for half sibs, and 0.5 for full sibs) and sigma2A is
the additive variance component (which is later used to calculate the
heritability). If I understand WinBUGS correctly, the program uses a precision
matrix as the second terms for the MVN distribution, i.e. the inverse of the
(co)variance matrix (A(sigma2A) in my example) similar to using the inverse of
the variance as the precision parameter in a simple normal distribution. If my
assumptions about the MVN distribution in WinBUGS are wrong please let me know.
I build my program so that WinBUGS computes the inverse of the (co)variance
matrix and then used that term as the precision matrix for the MVN
distribution. My matrices are fairly big (225 x 225) and I am suspecting that
maybe WinBUGS has trouble with inverting such big matrices and it is why the
computation step never ends. Does anyone encountered such problem before or has
any suggestions to help me solve my problem? Also, if anyone has implemented
the animal model in WinBUGS before and has any suggestions, tips, or any
comments that would be helpful I would greatly appreciate.
Thanks in advance,
Julie Marcil
Julie Marcil (MSc)
e-mail: [log in to unmask]
Gulf Fisheries Centre
Moncton, NB, CANADA
-------------------------------------------------------------------
This list is for discussion of modelling issues and the BUGS software.
For help with crashes and error messages, first mail [log in to unmask]
To mail the BUGS list, mail to [log in to unmask]
Before mailing, please check the archive at www.jiscmail.ac.uk/lists/bugs.html
Please do not mail attachments to the list.
To leave the BUGS list, send LEAVE BUGS to [log in to unmask]
If this fails, mail [log in to unmask], NOT the whole list
|