Using predicted gene expression from PrediXcan

[Lea Lemler]

Dear Haky, Dear Festus, Dear all,

After successfully predicting gene expression for our datasets (with PrediXcan) in the last months and doing different analysis with it, I have started to use the predicted gene expression for Ordinal Logistic Regression and Multinomial Logistic Regression. It turns out that in contrast to the other models (e.g. Cox, Fine & Grey, simple Logistic Regression) these models seem not to work well with it. I keep getting p-values of 0 for a large number of genes, which is not realistic. I do have quite a few genes with  predicted gene expression = 0. I don't know if this could have an impact maybe. However, I don't understand then why all other models work well with it but these two models seem not to function with it.

Have you encountered similar problems or have any idea why this might be the case? Or is there anything I could do?

Thank you very much & Best Regards,

Lea 

[Hae Kyung Im]

Hi Lea

not sure why things would be different with this approach. You may want to plot the predicted expression vs your phenotype, maybe boxplots by category for a few suspicious gene/trait associations.

Hakyk

--
You received this message because you are subscribed to the Google Groups "PrediXcan/MetaXcan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to predixcanmetax...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/predixcanmetaxcan/f71bd027-fcef-4837-b06e-fa83e8dfacdan%40googlegroups.com.

--

Hae Kyung Im, PhD - 任慧耕 - 임혜경 - She/her
Assistant Professor, Section of Genetic Medicine
The University of Chicago
Member of Committee on Genetics, Genomics & Systems Biology
Center for Translational Data Science

Computing Environment and Life Sciences Directorate, Argonne National Laboratory
5841 S. Maryland Ave. Chicago, IL 60637, USA| N412
Phone:  773.702.3898 | FAX: 773.702.2567
Email: ha...@uchicago.edu | Twitter: @hakyim
http://hakyimlab.org | https://github.com/hakyimlab | https://orcid.org/0000-0003-0333-5685
http://scholar.google.com/citations?user=1QD4sIcAAAAJ
Directions to my office https://tinyurl.com/hki-office