Entering edit mode
TB18
•
0
@tb18-15216
Last seen 6.3 years ago
Dear all,
I am running GOseq on a custom annotated dataset. I am getting a nicely looking output: list of enriched and depleted GOterms for my different groups, just as I want. But after calculating pwf I get the following warning message:
> pwf=nullp(cat_genes_vec,bias.data=lengths_features_with_GO) Warning message: In newton(lsp = lsp, X = G$X, y = G$y, Eb = G$Eb, UrS = G$UrS, L = G$L, : Fitting terminated with step failure - check results carefully
But pwf is calculated:
> head(pwf) DEgenes bias.data pwf 1 0 243 0.001000788 2 0 560 0.001000780 3 0 262 0.001000788 4 0 1176 0.001000628 5 0 248 0.001000788 6 0 291 0.001000788
Further, the plot from nullp is just a flat line on 0 (or more likely 0.001) hard to tell. I cannot clarify this. I think the warning message and plot result are related; I do get an output when I continue.
Can someone help me understanding this?
Thanks in advance!
Hi there. I am getting the same warning message while executing broadenrich() for GO and KEGG pathway enrichment.