removing outlier genes

[aelhossiny]

I have read in your documentation that

The new.prism function removes any genes in the mixture greater than 1% in more than 10% of the mixture samples by default. This threshold is very lenient, and usually not does not filter out a large number of genes.

I am using BayesPrism to deconvolve Visium spatial data. I have been facing a problem that this leads to removal of some of the hallmark genes that are known for certain cell types in my dataset. Is there a problem setting this to 0 or decreasing the thresholds to be more lenient?

Thanks

[tinyi]

Thank you for your interest in our methods.

This shouldn't be a problem. Based on my experience, Visium data typically has fewer outliers than bulk RNA-seq.