Hi Belinda,
Thank you for your work on the speckle package. It's been a great tool in our analysis of cell-type proportion data.
We’re currently using the propeller.anova function to compare cell-type proportions across conditions. However, we often need to:
-
Perform post hoc pairwise comparisons (e.g., Tukey or Dunn-style tests) to determine which groups differ significantly following a significant ANOVA, and
-
Visualize results with summary statistics like standard error of the mean (SEM) to aid in plotting and interpretation.
Would it be possible to implement support for:
Post hoc tests with p-values following a significant ANOVA,
Output of group-wise SEMs (or an option to compute them from the fitted model),
Or alternatively, guidance on how to extract these values from the existing model output?
Having these features would streamline interpretation and visualization, especially for multi-group comparisons in complex datasets.
Thanks again for the excellent package, and I’d be happy to help test or clarify any part of this request.
Best regards,
Hi Belinda,
Thank you for your work on the speckle package. It's been a great tool in our analysis of cell-type proportion data.
We’re currently using the propeller.anova function to compare cell-type proportions across conditions. However, we often need to:
Perform post hoc pairwise comparisons (e.g., Tukey or Dunn-style tests) to determine which groups differ significantly following a significant ANOVA, and
Visualize results with summary statistics like standard error of the mean (SEM) to aid in plotting and interpretation.
Would it be possible to implement support for:
Post hoc tests with p-values following a significant ANOVA,
Output of group-wise SEMs (or an option to compute them from the fitted model),
Or alternatively, guidance on how to extract these values from the existing model output?
Having these features would streamline interpretation and visualization, especially for multi-group comparisons in complex datasets.
Thanks again for the excellent package, and I’d be happy to help test or clarify any part of this request.
Best regards,