I noticed that many genes in a dataset (different from the one provided) have a high number of zeros. I'm talking about genes with zeros in almost all samples. This makes the data quite sparse.
I'm curious about a couple of things and would love some guidance:
How does HYFA handle these kinds of high-zero columns? Is there a specific way it differentiates true zeros from missing data?
Any tips on preprocessing or tweaking model settings in HYFA to better handle this sparsity?
Should I consider any data adjustments before feeding it into HYFA? I was thinking at the minimum removing all genes that have all values as zero.
I noticed that many genes in a dataset (different from the one provided) have a high number of zeros. I'm talking about genes with zeros in almost all samples. This makes the data quite sparse.
I'm curious about a couple of things and would love some guidance:
Thanks for any tips or advice you can share!