[GenABEL-dev] Interesting idea by Gulnara Svischeva

[L.C. Karssen]

On di, 2011-10-11 at 22:51 +0200, Yury Aulchenko wrote:
> 

> 
> This was indeed a major step -- with multiple improvements after that
> -- and now 'polygenic' can estimate the model for say 3k people in few
> minutes (in contrast with few hours / days? in old implementation).
> This development is thanks to Gulya's brilliant decomposition idea,
> comments of other people about it, and -- in some part -- my coding. 
> 

I've attached the results of a quick comparison (only one calculation
per data point) similar to the one you sent to this list on 2011-05-20.
I'm a bit surprised by the fact that the computation seems to flatten at
the end for all methods. After we have genotyped some more samples I'll
try to add another data point.

> 
> > Has there been any comparison of poly_eigen vs. the old method (in
> > terms
> > of variance, lambda etc.? 
> 
> 
> It is 100% identical (there is an GenA's RUnit test running 'old' vs
> 'new', and it passes checkEquals with tolerance of 1e-6 :) ). No
> approximation, just speed-up and still full ML. Cool, he?

Definitely cool! I noticed the equality as well and since I hadn't
looked at the math behind it I was surprised (your comparison of old
plogenic() with polygenic_hglm() does show differences), and wondered if
I made a mistake. Glad to hear that is not the case.


Lennart.


> 
> 
> best wishes,
> Yurii
> 
> 
> -------------------------------------------------------
> Yurii Aulchenko, PhD, Dr. Habil.
> Independent researcher and consultant
> yurii [dot] aulchenko [at] gmail [dot] com
> 
> 

-- 
-----------------------------------------------
L.C. Karssen
Erasmus MC
Department of Epidemiology
Room Ee2224

Postbus 2040
3000 CA Rotterdam
The Netherlands

phone: +31-10-7044217
fax: +31-10-7044657
email: l.karssen at erasmusmc.nl
GPG key ID: 0E1D39E3
-----------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: poly_scaling.pdf
Type: application/pdf
Size: 18095 bytes
Desc: not available
URL: <http://lists.r-forge.r-project.org/pipermail/genabel-devel/attachments/20111012/fa833929/attachment-0002.pdf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: poly_scaling_log.pdf
Type: application/pdf
Size: 17801 bytes
Desc: not available
URL: <http://lists.r-forge.r-project.org/pipermail/genabel-devel/attachments/20111012/fa833929/attachment-0003.pdf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://lists.r-forge.r-project.org/pipermail/genabel-devel/attachments/20111012/fa833929/attachment-0001.pgp>