Integration in single cell RNAseq: What does it mean?

Question

[synat.keam]

Dear Fellows,

Hope you are well. I am a very newbie in single cell RNAseq and I am learning to analyse the data in our lab. I am wondering whether you could explain the purpose of integration in a simple description.

Scenario: Say, we have three experimental groups (three sample per group), Group 1 (control), Group 2 (treatment: Drug A), Group 3 (treatment: Drug C). From what I read and understood, integration means combining cells from each sample together into a single sample of group 1, group 2 and group 3 rather than combining cells from group 1, 2 and 3 together? Kindly help explain!

Can we generate the percentage of cell types from each sample from single cell data, so we can do comparison between group like flow cytometry? or compare gene expression of each cell type between group.. I am looking forward to hearing from you soon.

Regards,

Synat,

Answer 1

[ATpoint]

Please go through: https://bioconductor.org/books/release/OSCA/

It covers all relevant basics and a lot more.

[synat.keam]

Thank you ATpoint. really informative. Do you also have a tutorial for bulk and single TCR seq. Regards,

Synat

[Chris]

Thank you! Is this section for integration? http://bioconductor.org/books/3.17/OSCA.multisample/human-pbmcs-10x-genomics.html#data-integration

It is different compare with integration with seurat https://satijalab.org/seurat/articles/integration_introduction

Answer 2

[bk11]

Also, please go through https://www.sc-best-practices.org/cellular_structure/integration.html

It covers why integration is needed, what methods can be used for this purpose and lot more.

[synat.keam]

Dear bk11,

Thank you very much. seems like it is python based. I really like it and will go through the tutorial. Do you also have a book or tutorial for TCRseq from beginning to advanced.

Regards,

Synat

[bk11]

Please see the link https://github.com/crazyhottommy/TCR-BCR-seq-analysis/blob/master/README.md. It has lot to offer. I had good luck with MiXCR before.

https://mixcr.com/