Question: What Is Benjamini
0
gravatar for mtyler.jason
14 months ago by
mtyler.jason10
mtyler.jason10 wrote:

I was doing some gene expression analysis using this tool http://david.abcc.ncifcrf.gov/summary.jsp. However, I have a confusion about what benjamini is. I fed it some gene list and it gave me some potential pathways the genes belong to(KEGG pathways). However, I have a confusion what this column called benjamini is. What does it denote? Any suggestions?

ADD COMMENTlink modified 14 months ago by Alex Reynolds6.2k • written 14 months ago by mtyler.jason10
9
gravatar for Woa
14 months ago by
Woa2.1k
United States
Woa2.1k wrote:

The column you're referring to contains 'corrected' p-value . The p-value is corrected for multiple hypothesis testing using the Benjamini-Hochberg method. There are other methods for such correction, the "Bonferroni-method" is one of them.

ADD COMMENTlink modified 14 months ago • written 14 months ago by Woa2.1k
1

Correct. Here's some useful background information. And here's a useful how-to which explains DAVID clearly, including the Benjamini column.

ADD REPLYlink modified 14 months ago • written 14 months ago by Neilfws41k

@Neilfws. Thanks for the information. However, I only know about t tests and pvalues. I don't have much knowledge about false discovery rate and all. The wiki article is a bit too technical for me. Can you guys give me something that can help me get better idea?

ADD REPLYlink written 14 months ago by mtyler.jason10
2

imagine that an event has 10% of chance of occurring randomly, now if you were to repeat the experiment twice then it is obvious that the chance of observing the event has to be larger than 10%,

The easiest way to estimate it is to compute the chance of not seeing the event in both trials .9*.9 = 0.81 thus the chance of seeing at least one event is then 1-0.81 = 0.19 = 19%. So even though you started with 10% chance, the corrected probability of seeing the event is 19%

ADD REPLYlink written 14 months ago by Istvan Albert ♦♦ 39k
Please log in to add an answer.

Help
Access
  • RSS
  • Stats
  • API

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.0.0
Traffic: 575 users visited in the last hour