Entering edit mode
6.4 years ago
hkarakurt
▴
180
Hello, I am trying to analyze a big dataset via student t-test. Data set includes two different data sets coded as GSE18489 and GSE31068 (they can be used together as authors told me). When I use ttest2 function of Matlab, I have about 5200 significantly changed genes. When I use Benjamini Hochberg correction via mafdr command, I have about 5600 significantly changed genes.
What can be the main reason of this result?
Thank you.
Seems a bit suspicious... are you looking at the results in Excel, Excel for Mac, or Open/Libre Office? Either of these may have distorted the results. Also, if you're in a 'Latin' nation, including Italy, Spain, Portugal, all of Latin America and Central America, and some African nations, then a comma will be the decimal specifier and not a dot.
If you're using R, be sure also that scientific notation is not causing an issue by setting
options(scipen=999)just before you generate P values. This sets decimal places to 999.Also specify the exact command that you use when correcting for P values.
Without seeing the p-values and knowing what you've done, it's hard to say precisely.