Your data analysis solution
Name is required.
Email address is required.
Invalid email address
Answer is required.
Exceeding max length of 5KB

ANOVA - Multiple comparisons

By Ida Rosenkrands | Apr 01, 2016 05:24PM CEST

I have analysed my proteomics dataset of 635 proteins with 3 replicates of 3 different growth conditions (RB, EB-Hela, ARB) for a bacterium. I have analysed the data by Differential expression ANOVA analysis with Benjamini-Hochberg correction, p<0.05 was considered significant.

I have two questions:

1) When you do multiple comparisons, how do you determine which growth conditions are significantly different? It was my understanding that they were significantly different if they had a different letter in the results file (a, b or c), but what if they have the letters ab, will it then significantly different from a and b or? See the example below:

Features p-value Significant RB EB-HeLa ARB
CT_320 0,001 Yes 8,377 (c) 7,797 (a) 8,145 (b)
CT_376 0,019 Yes 7,747 (ab) 8,098 (b) 7,397 (a)

2) Is it possible to calculate adjusted p values also for the individual comparisons, e.g. for the RB- versus ARB comparison for CT_320 above? If so, are these p values also changed after Benjamini-Hochberg correction compared to the analysis with no correction?

Thanks in advance for your reply,





By Efthalia | Apr 01, 2016 06:43PM CEST | XLSTAT Agent

Dear Ida,

If two categories share a common letter, then they do not differ significantly.
You can also verify this by the p-values (Pr > Diff) displayed in one of the tables for each comparison. These p-values are adjusted using the Benjamini-Hochberg correction.
Here is a tutorial that might be useful for you


This question has received the maximum number of answers.

Contact Us

Email Sales

Contact our technical support team:
seconds ago
a minute ago
minutes ago
an hour ago
hours ago
a day ago
days ago
Invalid characters found