Gökçen Eraslan

@sjfleming Is there a GIST or repo url to use this code? Might take time to integrate into scanpy/anndata but people can benefit from the code if it already lives...

Guys we should just keep the layer info here in log1p: data.uns['log1p'] = {'base': base} like data.uns['log1p'][layer] = {'base': base}

Super cool find Isaac, I didn't know about it! It'd be great to support it somehow. Then we can make queries like remove intervals overlapping with genomic regions with low...

@ivirshup is there an elegant way to get subset of genes with `_get_obs_rep`? because in the current code there are things like: ```python _adata = adata.raw if use_raw else adata...

I guess this is not maintained any more. Maybe someone should help porting https://github.com/wujian16/Cornell-MOE to Python 3.

I need that a lot actually. One question, is it different than just `sc.pl.umap(adata, groups=adata.bulk_labels.cat.categories)`? Maybe a special keyword like groups='all' would be easy to use on the API side?

I'm also interested in this since I'll be analyzing some HTO data soon. As I wrote [here](https://github.com/theislab/scanpy/pull/797/files/8bcee13537d6353399f1722bac7f60bc943a482f#r335664372), I think we should also discuss the I/O and storage procedures for ADT/HTOs....

This is the categorical equivalent of the `sort_order` parameter basically, right? I like the idea.

I don't get an error but just the following warnings: ```pytb Performing cosine normalization... /home/gokcen/.miniconda3/lib/python3.8/site-packages/mnnpy/utils.py:14: NumbaWarning: Compilation is falling back to object mode WITH looplifting enabled because Function "l2_norm" failed...

I'll have time next week to look into this.