[R] Logistftest to select diagnostic genes

Lingsheng Dong dong_lsh at hotmail.com
Wed Jan 18 18:14:18 CET 2006


Hi, all,

Anyone has experience on Logistf package? I am using logistftest in Logistf 
package to selelct diagnosis genes. The result seems not the same as I 
expected.

I have 10 gene expression data for 27 tumor 1 and 11 tumor 0. I want to 
select the best one using Maximum likelihood ratio test in logistic 
regression model. This is the way my code works:
1.  Read in 10 genes as independent variables and tumor type (1 or 0) as 
dependent variable.
2   Fit in a 10 variable logistic model and calculate it's likelihood.
3   Loop through the 10 genes, each time take one gene out of the model and 
calculate the likelihood without the gene. And compare each new likelihood 
with the likelihood of  the  original original 10 variable model to get 
likelihood ratio test for each gene.

I guess gene 8 should have best discrimination power because it can totally 
seperate the two group tumors. But the test result shows the likelihood 
ratio for Ratio 8 is not the biggest.

Where am I wrong? The package is not good for this problem? But the package 
document says this package takes care of separation and small sample size by 
finite parameter estimates and profile penalized log likelihood. Or my 
assumption is wrong? I am totally lost.

Thank you very much for reading this email! Your suggestion or comments will 
be appreciated.



Lingsheng






The fear of the LORD is the beginning of wisdom, and knowledge of the Holy 
One is understanding.
--Proverbs 10:10




More information about the R-help mailing list