[GenABEL-dev] Interesting idea by Gulnara Svischeva

[Yurii Aulchenko]

Dear Lars,

I am testing your procedure now; results will follow soon.

Meanwhile Gulya has developed R code implementing her idea to compute
the polygenic model likelihood in an efficient way, and I introduced
this as a default option in GenABEL's 'polygenic' "private" version.
'polygenic'  finds maximum likelihood estimates (via numeric
maximization of the likelihood function).

Recently, I convinced Gulya that we run minimal risk, but will have
multiple benefits if we go into the open now; so her code is now
included in the GenABEL public developmental version (on RForge). Here
is the link to the code of the likelihood function (see second one,
'polylik_eigen')

https://r-forge.r-project.org/scm/viewvc.php/pkg/GenABEL/R/polylik.R?view=markup&root=genabel

and here is the link to 'polygenic' FYI

https://r-forge.r-project.org/scm/viewvc.php/pkg/GenABEL/R/polygenic.R?view=markup&root=genabel

Our tests demonstrated excellent agreement between 'old' and 'new' way
of computing the likelihood and getting to MLEs -- results are the
same (some really minor differences could be attributed to finite
machine precision). The 'new' procedure works way faster than the old
one.

best wishes,
Yurii

2011/5/18 Lars Rönnegård <lrn at du.se>:
> Hi,
>
> I attach an implementation of an idea proposed by Gulnara. It uses weighted
> least squares to compute the REML heritability estimate.
>
> The function is found in “REMLrotateFun.R” and some example code is found in
> “ExampleCode_REMLrotate.R” that uses the data in “RemlEx.RData”.
>
>
>
> If someone would be interested in trying it (and possibly also suggest
> improvements), I would appreciate it.
>
> It works for linear mixed models having a polygenic random effect and iid
> residuals. It seems to be very fast. The example I attach contains 680
> observations.
>
> Is this a new algorithm or has it been done already?
>
>
>
> Best wishes,
>
> Lars Rönnegård
>
> Dalarna University
>
> _______________________________________________
> genabel-devel mailing list
> genabel-devel at lists.r-forge.r-project.org
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/genabel-devel
>
>