Skip to content

Promoter Detection Class Labels in GUE Dataset #137

Open
@mprincipato

Description

@mprincipato

Hello!

In the GUE/prom datasets, does the class "0" mean the sequence has a promoter and "1" mean that it does not have a promoter? Is it the other way around, where "1" means a promoter is present and "0" means it isn't present?

(Context:

I'd like to see how DNABERT/DNABERT-2, fine-tuned for promoter detection using the GUE promoter detection datasets, performs on a new set of promoter detection test data.

To do that, I was hoping to run the GUE code locally, but with the test dataset replaced with my own test dataset.

However, to make my test dataset match the training data, I need the classes to be able to match.)

Thank you!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions