Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

pocket rank differently between website and App #5

Open
suncyd opened this issue Jan 6, 2023 · 1 comment
Open

pocket rank differently between website and App #5

suncyd opened this issue Jan 6, 2023 · 1 comment

Comments

@suncyd
Copy link

suncyd commented Jan 6, 2023

Hello, I would like to like to ask a question that I have used PrankApp and the website to predict the same protein. But I found the final rank is different. So if there is some explanation for this result. Thank you so much.

@rdk
Copy link

rdk commented Jan 9, 2023

@suncyd Just to confirm: I'm assuming you mean that there is a difference between the results produced by standalone P2Rank command line app and website prankweb.cz. Is that the case?

In this case I think the difference is due to the fact that the default model (i.e. trained random forest) in the standalone app is different than one that is used as default in the webappp. Webapp uses model that includes sequence conservation score among features/descriptos while standalone app uses model without conservation. If you disable "Use conservation" checkbox on the website you should get the same results as with the standalone app.

In general, model with conservation is slightly more precise.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants