I used limma package in R to estimate the fold change, but I don't know which probes are up fold changes and which down? Also I need to know how to annotate the deferentially expressed genes? Do I need a new package or is Limma enough?
When I use LIMMA, I always have it output a file of results for my complete dataset, and that will include columns of Log2 Fold Change estimates with appropriate signage (if you were using Log2 transformed data as input).
The code would be something like:
t = topTable(ebayes, number=nrow(e))
write.table(t[1:31099,], file = "data_all_data_LimmaOUT.txt", sep = "\t")
where the numbers 1:31099 are just the total number of items in your dataset (in this example for an Affy microarray, which you could always determine by just "dim(e)" where "e" is your expression Affy batch expression object).
The table will be written as a tab delimited text file with columns for the feature ID, the estimated Log2 FC for each contrast in your design, the average Log2 expression over all samples, the F statistic, the P-value and the FDR-adjusted P-value.
If you also use the "write.fit" function to output a table with the individual contrast p-values for all features in your dataset, it is pretty trivial to then pair up the two files, matched by feature ID for both fold change and contast-specific pValue
e.g. write.fit(ebayes, file = "data_allData_separate_BH_adj.txt", adjust="BH", method="separate", sep="\t")
Thanks a lot for your help i will try using annotate and i noticed positive and negative values in my result thanks again
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools