search

YangLabHKUST / SpatialScope

A unified approach for integrating spatial and single-cell transcriptomics data by leveraging deep generative models
https://spatialscope-tutorial.readthedocs.io/en/latest/
GNU General Public License v3.0
50 stars 6 forks source link

scRNA-seq reference integration #8

Closed ChenJin2020 closed 7 months ago

ChenJin2020 commented 8 months ago

Hi,

I am currently conducting research on ovarian tissue using 10x single-cell RNA sequencing (scRNA-seq) datasets. However, these datasets do not include oocytes, which are crucial for my study. To address this, I plan to integrate my 10x scRNA-seq datasets with publicly available scRNA-seq datasets of oocytes, which were generated using the SMART-seq2 method. My question is about handling the differences in data resolution and depth between these two methods. Specifically, when I use the combined datasets as a reference in my research, should I downsample the SMART-seq2 oocyte datasets to match the resolution of the 10x scRNA-seq datasets? I'm concerned about potential biases or discrepancies in data interpretation due to the different sequencing depths and resolutions of these methods. Any guidance on this integration process would be greatly appreciated.

Thanks,

Chen

JiaShun-Xiao commented 8 months ago

Hi,

I am currently conducting research on ovarian tissue using 10x single-cell RNA sequencing (scRNA-seq) datasets. However, these datasets do not include oocytes, which are crucial for my study. To address this, I plan to integrate my 10x scRNA-seq datasets with publicly available scRNA-seq datasets of oocytes, which were generated using the SMART-seq2 method. My question is about handling the differences in data resolution and depth between these two methods. Specifically, when I use the combined datasets as a reference in my research, should I downsample the SMART-seq2 oocyte datasets to match the resolution of the 10x scRNA-seq datasets? I'm concerned about potential biases or discrepancies in data interpretation due to the different sequencing depths and resolutions of these methods. Any guidance on this integration process would be greatly appreciated.

Thanks,

Chen

Hi,

We have not yet attempted to integrate the 10x and SMART-seq2 single-cell datasets. But we think it is okay to do this, and ideally the sequencing depth issue can be solved through normalize_total preprocessing. You could try using the integrated single cell reference for "cell type identification" and see if the oocytes in your spatial data are correctly identified. If this step is successful, the integrated single-cell reference is totally fine.

Thanks, Jiashun

ChenJin2020 commented 7 months ago

Thank you