Question: Enrichment analysis for class comparison
1
3.6 years ago by
Bnf83150
Bnf83150 wrote:

Hi Guys, sorry if the question might sound stupid but I'm not sure about how to set a statistical test. Basically I have 2000 breast cancer patients that had surgery. Of them, 1000 had additional tumours (not secondary but primary tumours) divided according to the following classes.

``````ADDITIONAL
Class A      50
Class B      250
Class C      450
Class D      250
``````

Total: 1000

The 1000 patients are further divided in pre-post surgery as follows:

``````PRE
Class A       20
Class B      150
Class C      120
Class D       75

POST
Class A       30
Class B      100
Class C      330
Class D      175
``````

Now I would like to compare the dataset PRE with the dataset ADDITIONAL in order to see if there's some enrichment as well as the dataset POST with the dataset ADDITIONAL for the same reason. I already have compared the PRE-POST between them. I suppose I have to perform a Fisher test but I' m a little bit confused on how to write the contingency table. Can anyone help me please?

Thank you in advance

Best regards

B

fisher R • 628 views
modified 3.6 years ago by WouterDeCoster45k • written 3.6 years ago by Bnf83150

I think you should change your post from Job to Question

I just did that ;-)

Hi,

column PRE + column POST = column ADDITIONAL

So, I am unsure whether it makes sense to compare one PRE or POST column to the margin column.

My two cents.

ADD REPLYlink modified 3.6 years ago • written 3.6 years ago by SamGG20

Hi, if I well understand you are showing only half of the data. You said that you have 2000 patients. So 1000 are in the ADDITIONAL group and 1000 are in another group NO_ADDITIONAL. For the NO_ADDITIONAL you should create again the PRE and POST tables and compare them. Also, it would help if you further detail what kind of enrichment are you looking for. Enrichment in categories of PRE compared to POST or enrichment in categories of PRE in ADDITIONAL compared to PRE in NO_ADDITIONAL?