Count Matrix from Single cell RNA

Question

[David_emir]

Hi All,

We did Single Cell RNA data analysis from 25 diseased patients. The Vendor sent 25.MTX files to us. The files we received is as follows

    filtered_feature_bc_matrix
├── barcodes.tsv.gz
├── features.tsv.gz
└── matrix.mtx.gz
0 directories, 3 files

The data comes off the machine as bcl files and is converted to counted matrix files using 'cellranger mkfastq' and 'cellrangercount. My question is we have 25 different.mtx.gz files and we sequenced ~ 2000 cells/Sample.

1. Should we merge these 25.mtx.gz files to create ONE new Master Matrix file for further analysis? like Columns(2000*25) of cell barcodes and 32000 rows of gene name into one super Master file. like how we used to do in Bulk RNA seq data analysis.
2. Whats the best way to annotate Gene names?
3. Please let me know the best pipeline for scRNA data analysis - we are biologists with limited experience in coding.

Thanks a lot for helping us, Have a great day ahead.

Dave

[GenoMax]

Take a look at vignettes for Seurat from Satija lab here.

[David_emir]

Thanks, Genomax, I have gone through this, in this, they have not mentioned about merging samples. the example data is of 2,700 single cells that were sequenced on the Illumina NextSeq 500. What we have is 25 patients (Prostate cancer) data and 25 different matrix files of the same disease and each sample have got 2000 cells. Should I individually read the 25 different count matrix files and merge?

[GenoMax]

It is my understanding that you will need to analyze 25 samples individually to make sure they all pass QC. You can then start integrating the samples to do comparative analysis. Someone with more experience with scRNA analysis will be along to provide specific pointers. If you have never done this type of analysis then be ready to spend a substantial time/effort, as you learn.