filtering only differentially expressed genes

Question

[zizigolu]

Hi,

I used a tool to perform differential expression analysis on my microarray data sets,

The output is something like below

ProbesetID    logFC    AveExpr    t    P.Value    adj.P.Val    B
1769308_at    1.89394112596704    10.5871328585202    16.4749776727376    3.68669391559684e-10    2.6623530112e-08    13.7014616531283
1769311_at    -0.115715211915935    10.5999601204582    -1.12002479223092    0.282752344869474    0.382582735789368    -7.1722215976202
1769312_at    0.131966902231916    9.02503563963962    1.5364411821642    0.148147122010043    0.230066458270885    -6.6497072352879

Sorry, may someone please tell me how I can select only the probsets with a desire fold change or P.Value?

For example probesets more than 1

Thank you so much

Answer 1

[Sean Davis]

In R, take a look at the help for subset().

[zizigolu]

thank you,

but I am too weak in R and can't figure out easily what to type in R

[Sean Davis]

Did you read the help for subset by typing help("subset")? Notice that the help for R functions usually includes one or more examples at the end of the help page which you should read and type into R by hand. Also, a quick google search turns up: http://www.r-bloggers.com/subsetting-data/ and http://www.ats.ucla.edu/stat/r/modules/subsetting.htm among many others. If this doesn't help, then I would advise an online course in R before proceeding, as subsetting is a basic necessity for doing any work in R.

[svlachavas]

I assume the above output is from topTable, right?

for instance(assuming that my above notion is correct)--as you have a data.frame, let's called it dat, which has the output of topTable. Then, you could use something like the following:

dat2 <- dat[which(abs(dat$logFC) >cutoff1 & dat$adj.P.Val < cutoff2),] # in order to satisfy both cutoff1 and 2-where the two cutoffs depend on the user

Hope that helps,
Efstathios

[Sukhi Singh]

then first thing you spend your time on is R.

• Web Based Interactive Resources To Learn Programming And/Or Bioinformatics
• A few words for R beginners

Answer 2

[Antonio R. Franco]

If having problems with R, you can simply write the table with the data, and import it into excel. Filtering and ordering in excel is known for you, I believe..

[Sukhi Singh]

This can go in wrong direction, why not learn R?

[zizigolu]

actually you all right, but one month more I should leave Germany while I don't have any result to defend my dissertation. you can't imagine how much I am worry and concentrate on anything and only I am trying to collect some results...never I thought the final of my studying here would be like this... :(