Warning message when duplicated samples

[shikhanayar]

Hi, using scCOMP for the first time; the issue I'm running into is that for some reason cell_group and sample calling are not working appropriately from a Seurat object when running scCOMP.

Here is the code (contrasts have been pre-defined):

results1 <- seurat |> sccomp_glm( formula_composition = ~ 0 + disease_inflammation_response, contrasts = contrasts, .sample = sample, .cell_group = cell_type, bimodal_mean_variability_association = TRUE )

Here is the error message:

sccomp says: outlier identification first pass - step 1/3
Joining with by = join_by(sample)Joining with by = join_by(cell_group)error occurred during calling the sampler; sampling not done
error occurred during calling the sampler; sampling not done
error occurred during calling the sampler; sampling not done
error occurred during calling the sampler; sampling not done
error occurred during calling the sampler; sampling not done
error occurred during calling the sampler; sampling not done
here are whatever error messages were returned
[[1]]
Stan model 'glm_multi_beta_binomial' does not contain samples.

[[2]]
Stan model 'glm_multi_beta_binomial' does not contain samples.

[[3]]
Stan model 'glm_multi_beta_binomial' does not contain samples.

[[4]]
Stan model 'glm_multi_beta_binomial' does not contain samples.

[[5]]
Stan model 'glm_multi_beta_binomial' does not contain samples.

[[6]]
Stan model 'glm_multi_beta_binomial' does not contain samples.

Error in draws[, draws_colnames, drop = FALSE] :
no 'dimnames' attribute for array

Any help would be great.
Thanks!

[stemangiola]

Hello @shikhanayar,

can you send me a minimal object, anonymised if needed, so I can test it?

[shikhanayar]

Thanks for your reply @stemangiola . I was able to resolve the above issue (I think that there was an issue in my samples column).

[stemangiola]

Thanks for your reply @stemangiola . I was able to resolve the above issue (I think that there was an issue in my samples column).

Great, however, knowing your specific case would help me create a warning that would leave users less confused if this happens.

[shikhanayar]

For sure, I believe that some of the sample rows were duplicated, so when creating a dataframe of cell type and samples, there weren't complete distinct values. Follow-up question -- are there scenarios you have seen where 2D plots were not computed by plots<- plot_summary(results)?

[stemangiola]

are there scenarios you have seen where 2D plots were not computed by plots<- plot_summary(results)?

Can you be more elaborate more?

[shikhanayar]

Sorry, when I run the above function using the results dataframe, the only plots generated are boxplot and credible_intervals_1D. In your vignette, I also see the credible_intervals_2D plot but I'm not seeing that here.

[stemangiola]

I am refactoring the plots in a new release. 1d vs 2d depends on how you model the variability.

I will keep you posted.

[stemangiola]

Can you please reinstall from github and let me know?