Page 1 of 1

[Solved] Methodology for prediction equations of phenotypic

Posted: Fri May 20, 2011 7:35 pm
by Nicolas Guillemin
I'm working on a methodology for developing prediction equations of phenotypic data, based on SNPs, calculated with the lm function.
I saw that it's possible in the GenABEL tutorial.

I'm wondering if someone has experience in this domain, or if there is some interresting bibliography on it (I don't find publications with GenABEL on prediction model for phenotype based on SNPs) ? My interest is to demonstrate that this type of methodology, using lm and GenABEL, could be reliable for the project.


Re: Methodology for prediction equations of phenotypic data

Posted: Mon May 23, 2011 1:47 pm
by sumankundu
Dear Nicolas,
I am currently associated with comparing different simulation methods which have been used in (geneic) risk prediction studies and working with Cecile Janssens group in Erasmus MC, Rotterdam.
Following links may be usefull in your project.

1. PredictABEL: an R package for the assessment of geentic risk prediction model, developed by us.

2. Strengthening the reporting of genetic risk prediction studies: the GRIPS statement.

Best wishes,

Re: Methodology for prediction equations of phenotypic data

Posted: Thu May 26, 2011 1:55 am
by Nicola Pirastu

I'm not sure I completely understood what you're asking could you be a little more specific?
I don't think you can find any reference for using GenABEL to do linear regression with lm since lm is a function in the base package of R.

Anyways if you're trying to do something like:


so let's say your trait is qt1 from the srdta example data and you want to use rs18 rs29 and rs73 as predictors with lm, you could do something like:

Code: Select all


Does this answer to your question?

Hope I was of help.


Re: Methodology for prediction equations of phenotypic data

Posted: Mon Jun 13, 2011 9:13 pm
by Nicolas Guillemin
Thanks for your answers.

According scpecialized people, the GenABEL package is accurate for my project (modeling phenotypic data from SNPs).
So, I will use this package during my work