Hello,
I just started working with limma and so far followed the limma users-guide to recive a TestResults Object.
I've printed the results into a file using write.fit() and a question arose while looking at the data.
Say I was looking at differential expressed genes in samples X vs. Y.
From my understanding, the values in column Res.X-Y indicate significant differential expressed genes for given p-value and adjustment method, where for value 1, the gene is significantly more expressed in X etc..
Now, how do I sort those significant differentially expressed genes so my output ranges from most-significant to least-significant (within the as significant annotated ones).
Does it make sense to just sort by Coef.X-Y decreasing or p-value increasing?
Does it even make sense to treat as significant annotated genes differently?
I just started to deal with limma and (microarray data in general) during my current semester, so please go easy on me in case I missed something obvious.
Many thanks in advance!
Hey Alex,
I've checked out topTable() and it is indeed what I was looking for. However, what if I choose 'global' as a method for decideTests() ? From my understanding, the Significant Genes found by topTable() and decideTests() could vary in this case.
As topTable() ranks Genes by F-Statistics, could I simply sort the Genes in TestResults by F-statistics and thus get the same ranking as in topTable() ?
in topTable, you can also use the sort.by function
so assuming you were writing your code it would be something along the lines of: