search

AlexsLemonade / scpca-docs

User information about ScPCA processing
https://scpca.readthedocs.io/en/latest/
BSD 3-Clause "New" or "Revised" License
0 stars 1 forks source link

Addition of Multiplex FAQs #83

Closed allyhawkins closed 2 years ago

allyhawkins commented 2 years ago

Closes #77, #73, and #76. Stacked on #70.

I'm adding the three FAQs for multiplexed samples here, what is a multiplexed sample, what are est. demux cell counts, and why do multiplexed samples have est. demux cell counts. I decided to pull this together into one PR, because I thought it would make sense for reviewers to look at all of these together as they are all somewhat linked, especially the ones about estimated demux cell counts.

I also stacked this on #70 so that I could link to some of the information in the processing and sce file contents docs in answering the questions.

The one main question I have is if we want to explicitly say that we are reporting the cell counts using genetic demultiplexing? Right now as things are that's what we are using, but I know there was a concern that in the future that wouldn't always be the case. But it feels like something we probably want to include?

Additionally, is there anymore detail or information that I'm missing that should be included in these explanations?

As discussed in sprint planning I will request @jashapiro for a first round of review and then request a second reviewer who is less familiar to get their input.

allyhawkins commented 2 years ago

Thanks for taking a look at this @jashapiro! I would agree that the second two questions seemed a little repetitive and inter-related. What I did was combine them into one answer, focusing on how we are providing an estimate of cell counts before download but they are simply an estimate. Let me know if the updated answer has a better answer? Or if there was something else you had in mind? I tried to make the answer more focused.

I also added a new question for "Why are demultiplexed samples not available?" and explained that there is little consistency between the results of demultiplexing methods and that users can find the results in the SCE object and should make decisions on their own. I also reordered them so that this question appears first and then you have the question about what are estimated demux cell counts last. I thought this made sense since a lot of the reasoning behind estimated demux counts is because we aren't separating samples. I could repeat some of the same rationale about demultiplexing giving differing results in the "what are estimated demux cell counts" if we think that having it in both FAQs would be important. This should be ready for another look.

allyhawkins commented 2 years ago

@jashapiro I went ahead and added in a link to the notebook for now. I included a separate sentence with the link stating that we had performed some exploratory analysis comparing the different methods. Let me know what you think of that.

Alternatively, and probably outside the scope of this PR, we could do something similar to what we did with the FAQ comparing alevin-fry to Cell Ranger and create specific figures and link to those rather than the html file, if there's something in particular that you think would be helpful rather than the entire notebook. But I leave that discretion up to you.