1
$\begingroup$

 The blue dots represents enzymes and red dots represent non-enzymes

The score plot from the PCA is represented in the diagram. I performed the Mann Whitney U test for the Scores values for enzymes( blue dot in score plot) and non-enzymes ( red dot) for both PC1 and PC2. I found out that for both Pc1 and PC2, the enzymes and non-enzymes values were significant( both had p-value 0.00). Is there and way to find out by mann whitney U test ( by considering U value or any other values, whether PC1 or PC2 is more effective to seperate enzymes or non-enzymes?

$\endgroup$
4
$\begingroup$

Is there and way to find out by mann whitney U test ( by considering U value or any other values, whether PC1 or PC2 is more effective to seperate enzymes or non-enzymes?

Yes.

I don't know if this is really what you ought to do, but to answer your question, recall that the U-statistic is related to the AUC (Area Under the Receiver Operator Characteristic), which is a metric of binary "separation" over a continuous score. See here: wikipedia link.

$$AUC={U\over n_1\cdot n_2}$$

The higher the AUC, the bigger the separability of labels in the given continuous score.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.