Skip to content

Questions about normalization output and modeling on it #44

Description

@ccnawrocki

I am working on a project that utilizes some geomx data, and I have tried many different normalization methods. I am now trying out yours. My understanding is that I can model normalized expression with a linear mixed model for DE analysis. Can I use poisson threshold normalized data as the input data to a linear mixed model? I see that this package supports negative binomial linear mixed modeling for the counts, but could I use the normalized data with a custom function that uses lme4? Modeling with the assumption that the data is approximately normal speeds things up a lot.

Also, I am noticing that the poisson threshold normalized data contains negative values. Is this expected? Is the data z-score scaled? Typically, if log2(count+1) is performed, there should be no negative values. Am I missing something here?

Any help would be much appreciated. Thanks in advance for your time.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

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