Some doubts about the limma package
1
0
Entering edit mode
llkxiaolan • 0
@llkxiaolan-13767
Last seen 4.6 years ago

Recently I have been learning limma manual, as a white letter to read the manual completely read mumbo-jumbo, there are many places do not understand, there is no doubts about God, be grateful.

Such is the case ：

fit <- lmFit(eset.rma[present.probes,], design)

fit2 <- contrasts.fit(fit, contrast.matrix)

fit2 <- eBayes(fit2)

What do these lines mean? I know the third step is the Bayes test, but what does Bayes test do? And what's the function of the first two steps?

Finally, there are a few more questions about the Affy package ，I read the manual and felt that the main function of the Affy package was for chip pre processing .For example, restore the original image of the chip ,draw RNA degraded images to see the quality of the chip ,draw the bottom figure to see whether the chip data standardization ,then standardization using RMA methods .Finally, the limma package is used to analyze standardized data .Is my understanding correct, please?

Thanks

limma affy • 874 views
0
Entering edit mode

You need to read the limma papers, just search for 'limma linear models' in google scholar or anything like that.

Limma uses normalised data, like quantile normalised, you should be able to normalise within the package.

0
Entering edit mode

First of all, thank you very much for your reply. Secondly, I've read some of the limma manual. But as a Chinese, I'm a little tired of understanding

0
Entering edit mode

No the primary research papers are what you need to read buddy.

1
Entering edit mode
Aaron Lun ★ 28k
@alun
Last seen 5 hours ago
The city by the bay

I'm afraid I haven't seen God on the support site, so you'll have to make do with the rest of us.

?lmFit # fits a linear model
?eBayes # performs empirical Bayes shrinkage
?contrasts.fit # "refits" the model for specified contrasts


There is no "Bayes test". There is only empirical Bayes shrinkage of the variance estimates, or a moderated t-test (or F-test, for ANOVA-like comparisons) using those shrunken variance estimates.

As for your final question: limma provides some steps for background correction and normalization. (See Section 6 of the user's guide for details.) However, it can also accept pre-processed data from other sources, all you need is an expression matrix.

0
Entering edit mode

Thank you very much. After reading a book on the application of R in biology, I know how to do it