[R-sig-genetics] Bayesian network

[Lu, Kevin]

Hi Thibaut,

Thanks for your suggestion, and sorry for my late reply. The reason for
my try of the Bayesian network to find the markov blanket of my
phenotype is about the high order of gene-gene interaction. In the
outcome of Bayesian network, I can check these high order interactions
with causal suggestion. After so many trys and errors, I realize that
deal package can not deal too many variables due to limitation of the
memory (even using the computer cluster). I am now trying to use other
Bayesian network package, hopefully to reach my goal.

Best wishes,

Kevin

-----Original Message-----
From: Jombart, Thibaut [mailto:t.jombart at imperial.ac.uk] 
Sent: woensdag 10 juni 2009 13:14
To: Lu, Kevin; r-sig-genetics at r-project.org
Subject: RE: Bayesian network

Dear Kevin, 

I am admittedly not into Bayesian networks, but in any case the 10 SNPs
limitation sounds a bit surprising. However, the error you encounter is
quite package-specific. I think you may want to ask your question
directly to deal's maintainer, in case you would not find an answer
here. 

I don't know of any package that would implement Bayesian networks
specifically for genetic data.

Best regards,

Thibaut.

________________________________________
From: r-sig-genetics-bounces at r-project.org
[r-sig-genetics-bounces at r-project.org] On Behalf Of Lu, Kevin
[kevin.lu at wur.nl]
Sent: 08 June 2009 17:38
To: r-sig-genetics at r-project.org
Subject: [R-sig-genetics] Bayesian network

Dear all,

I am now trying to use Bayesian network to analyze my SNP data,
associating 20 SNPs with one phenotype (plasma cholesterol levels). I
want to use deal package to find the markov plane of my phenotype. I
have 3575 records. After trying the deal package, I can not go through
coomand----jointprior (). It is said "Error in array(1, Dim) : 'dim'
specifies too large an array".

I reduced the number of SNP to 10 and I can continue.  When I try
"result<-heuristic(my.data1.nw,my.data1,my.data1.prior,restart=2,degree=
10,trace=TRUE)". It was said that "[Heuristic [Autosearch Error in
phi[[y]] : attempt to select less than one element". So, I can not get
the results.

I hope someone could help me with it. Am I using the right package?

Thanks in advance,

Kevin



        [[alternative HTML version deleted]]

_______________________________________________
R-sig-genetics mailing list
R-sig-genetics at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-genetics