Update CITE-seq section of getting started docs to include AnnData

[allyhawkins]

Closes #164

Here I continued with the progress I had started in #162 and modified the CITE-seq section to both refer to only the processed objects and add in AnnData objects. Generally, I follow the format I started in #162 to include an overview of the step or item (like filtering and QC) together but then show two separate commands for R and python.

• Any duplicate commands for the processed and filtered objects have been condensed so that we only show how to work with the processed objects. This is what we did for the main section.
• I condensed both accessing the ADT data and getting the raw/normalized counts into one code chunk for R and one for python. I'm curious what others think of this, but I think I like this better. I don't think we need a whole separate chunk for reading in a HDF5 file when we already show them how to do it for the main RNA data.
• I also took a look at the multiplexed section and added a note that multiplexed libraries are only available as SCE objects. Let me know if I should add more there, but I think that note should be enough.

Please let me know if there's any places you think I missed.

[allyhawkins]

I also wonder if we want to provide some information here on converting to AnnData? Or maybe link to a different source, like zellkonverter docs?

Okay, I did add a note about converting and a link. But I didn't go into the weeds of separating out objects or that you should demultiplex in . I think that's okay, but let me know what you think.

I'm thinking about this sentence, because there are really two separate things we are trying to convey. Maybe we need to expand this a little to get these related but separate points across?

Regarding this comment, I actually don't want to talk about doing your own normalization from the filtered object, since we are encouraging people to use the processed objects. So I just removed the word "normalization" from that statement.

[sjspielman]

But I didn't go into the weeds of separating out objects or that you should demultiplex in

agreed, no weeds!

Regarding this comment, I actually don't want to talk about doing your own normalization from the filtered object, since we are encouraging people to use the processed objects. So I just removed the word "normalization" from that statement.

👍