[BioC] SAM slots ?

Giulio Di Giovanni perimessaggini at hotmail.com
Tue Jul 26 13:55:17 CEST 2005


Hi to all,


I'm using sam function in siggenes package. Everything seems to be perfect. 
But when I try to summarize the results (for example, running the siggenes 
vignette) I'm not able to access directly the slots with the results.
In the help file if  I'm not wrong is written that the ouput of the sam 
function has to be a list, with specified slot... How I can access them ?

(first: I found some output differences with the siggenes vignette...there, 
in the vignette, slots are reported in the output).

Here's what I obtain running the vignette (BioC 1.6, R 2.1.0, WinXP):

>sam.out <- sam(golub, golub.cl, rand = 123, gene.names = golub.gnames[, 3])

>summary(sam.out)

SAM Analysis for the Two-Class Unpaired Case Assuming Unequal Variances

s0 = 0.0584  (The 0 % quantile of the s values.)

Number of permutations: 100

MEAN number of falsely called genes is computed.

   Delta  p0   False Called   FDR cutlow cutup   j2   j1
1    0.1 0.5 2424.77   2739 0.443 -0.177 0.228 1434 1747
2    0.7 0.5  262.21   1248 0.105 -1.264 1.438  737 2541
3    1.3 0.5   12.11    507 0.012 -2.299 2.488  311 2856
4    1.8 0.5    0.74    210 0.002 -3.154 3.311  134 2976
5    2.4 0.5    0.01     76 0.000 -4.157 4.259   44 3020
6    3.0 0.5    0.00     15 0.000 -5.577 5.139    4 3041
7    3.6 0.5    0.00      5 0.000   -Inf 5.971    0 3047
8    4.1 0.5    0.00      2 0.000   -Inf 7.965    0 3050
9    4.7 0.5    0.00      2 0.000   -Inf 7.965    0 3050
10   5.3 0.5    0.00      0 0.000   -Inf   Inf    0 3052

>

That looks perfect.
But, for example, the siggenes vignette says:

"the general information about the set of significant genes [can be 
obtained]  by
>sum.sam.out$mat.fdr
NULL
"

That's what I have: NULL

???

And the same is for all the specified (in the help file) slots...
How can I have the FC list, the gene names, and all the results I need to 
continue my analysis?
Where are the slots ?
Well, what I'm doing wrong ?
Any help will be appreciated !!!!

Thanks a lot

Giulio



More information about the Bioconductor mailing list