[GenABEL-dev] faster polygenic

[Xia Shen]

Hi Yurii,

I just noticed that in the example for testing polygenic effect in polygenic_hglm, I wrote:

pchisq(S, 1, lower.tail = FALSE)*2

to be the p-value, but I think it should be 

pchisq(S, 1, lower.tail = FALSE)/2

instead, since the same statistic value would result in a smaller p-value in the mixture distribution compared to a chi-square(1) distribution. So it needs to be revised.

Xia

On Apr 18, 2011, at 11:16 AM, Yurii Aulchenko wrote:

> Many thanks, Xia, I have your changes committed to the devel-version.
> 
> Should update quite a few posts on forum when 1.6-6 released -- with
> polygenic_hglm many answers become quite trivial!
> 
> Only change I did
> 
> #' @keywords hglm, kinship, polygenic
> 
> to
> 
> #' @keywords htest
> 
> My understanding is that 'keywords' must one of the valid keywords
> which can be viewed with
> 
> library(gtools)
> keywords()
> 
> best wishes,
> Yurii
> 
> 
> On Sun, Apr 17, 2011 at 3:49 PM, Xia Shen <xia.shen at lcb.uu.se> wrote:
>> Hi Yurii,
>> 
>> Attached is an updated version of polygenic_hglm.
>> 
>> - The reference list is now in a chronological order.
>> - Default convergence criteria is now 1e-6. In the documentation of @param conv, we say for great precision, use 1e-8.
>> - An example is provided showing how one can test the fixed effects parameters and how the standard errors can be extracted.
>> 
>> Cheers
>> 
>> Xia