Question: fdr score for degs
gravatar for nia
10 months ago by
nia0 wrote:

hello fellows!!!!!!!!!!!

i have 10 samples from Geo data set  and there was 1067 probe ids its showed (pvalue<0.05) was set  after applying limma code (fdr < 0.5) was set , then 294 probeids with genes symbol and with all the other relevant info. was retrieved from 294 113 was upreg and 181 was down reg.

q1>Is it fine to choose fdr<0.5 because fdr<0.05 ,0.01,0.1 shows no result thats why i choose 0.5 is it acceptable or not this is a  question.

q2>some probe ids show entrez id and all the other info. but they don’t show gene symbol so should i consider it or not.

ADD COMMENTlink modified 10 months ago by svlachavas570 • written 10 months ago by nia0
gravatar for svlachavas
10 months ago by
Greece/Athens/National Hellenic Research Foundation
svlachavas570 wrote:

Dear Nia,

your question if using a specific fdr cutoff is very general, and has not a specific answer-using a specific FDR cutoff-lets say 0.05- is generally arbitary-conventional and mostly there is no "consensus" about the "best fdr", as it simply depends mostly on what conclusions you want to draw from your experiment, your experimental design, etc. Just keep in mind, that increasing the threshold, increases also the percent of putative false positives in your DE list. 

Regarding the annotation issue, one would argue that you could simply discard these probesets which do not match to any gene symbol, but again it is up to you for further verification.



ADD COMMENTlink written 10 months ago by svlachavas570

thankyou so much i got the point

ADD REPLYlink written 10 months ago by nia0
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.2.0
Traffic: 392 users visited in the last hour