search

immunomind / immunarch

🧬 Immunarch: an R Package for Fast and Painless Exploration of Single-cell and Bulk T-cell/Antibody Immune Repertoires
https://immunarch.com
Apache License 2.0
306 stars 65 forks source link

True Diversity, D50 vs shannon entropy #104

Open talban14 opened 4 years ago

talban14 commented 4 years ago

Hi, I'm newer to TCR analysis but most of the analyses that I see show Richness and Evenness in their analysis. I see that Immunarch can do Richness very easily with the chao1 but is there a way to show evenness? Also along the same line is there a way to show the shannon entropy between groups? This may already be shown in other methods or maybe there's another method that's better than evenness or shannon entropy, and if so can you please explain for my understanding of how to present the data output?

Thank you!

best, Tyler

EugeneRumynskiy commented 4 years ago

Hello @talban14 ! Thank you for reaching out. As for the questions, could you elaborate on them a bit?

  1. ...show Richness and Evenness ... It would be awesome to have a link to the publication.
  2. Could you give a data sample or more detailed explanation on groups for the shannon entropy. What is the data format? How big is the group? What is the grouping criterion?
  3. What is the problem with data output presenting? Could you describe the problem and desired outcome?

Also, it would be awesome to contact with you via email, could you send me a note at erumynskiy@immunomind.io :)

best, Eugene

talban14 commented 4 years ago

Hi Eugene, thanks for the response!

  1. attached is a link the the paper i'm refering to where they show richness vs evenness https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5685550/

  2. I'm not sure how best to share the data but I think it should be pretty standard output from MiXCR. We first used MiXCR on rna-sequenced tumor samples from two groups (n=17) in group A and (n=17) in group B and then I used the immunarch package to load the clone files along with metadata file. Everything worked and read into immunarch just fine and the data is easily graphed out.

  3. For me not being familiar with the statistical terminology that's used to describe TCR diversity I was trying to play it safe by showing shannon entropy (which I think is represented in some of your statistical outputs?). If shannon entropy is not included in your package maybe these other statical methods are better and if so can you clarify?

Sorry for any confusion, your package is great and works very well i'm just trying to make sure I understand the outputs correctly.

best, Tyler

aislinnjennings commented 2 years ago

Hi.

I was wondering also if this issue has been addressed?

Thanks. Ais

alxndravc commented 2 years ago

Hello, are there any updates on this? I just thought of the exact same question... Thanks in advance!

MVolobueva commented 2 years ago

Hi, @alxndravc!

The functionality of Immunarch you are asking about is not yet available. But we are working on it;) It will appear in the future.

We will get back to you with any updates.

Enjoy your week, Maria