Questions about Alignment and downstream workflow for scRNA-Seq Data

Question

[rajeev.vikram]

Hi,

I am new to single cell (SC) RNA-Seq and have a rather naive q regarding the alignment of single cell fast q files. As, there are hundreds/thousands of fast q files corresponding to their respective SCs, How do we align all these fastq files in parallel? should one align a few fastq files at a time or all these (hundreds/thousands of fast q files) all at once in some aligner? (Hisat/STAR). What is the preferred method? Is there any online instruction available regarding this (I didn't find anything addressing this specifically, that why I think its probably too naive) I am fairly conversant with bulk sequencing methods so any link to a fairly basic but recent workflow would be a great help.

PS: I am working with about 4k publicly available cancer single cell fast q files.

Thanks

[swbarnes2]

Are you sure your technique will result in a separate file for each cell? Many techniques do not.

[rajeev.vikram]

I was talking from a bulk-seq perspective, where we have one(or two) fastq files per sample. I am not really aware of how scRNA-Seq works, probably in batches, but how, thats why I asked the question. An example would be great.

Answer 1

[Devon Ryan]

STAR is actually pretty convenient, in that you can have it load the index into memory and leave it there for multiple alignment runs. That ends up saving a couple minutes per run, which, given the low sequencing depth of individual cells, can add up to a serious performance increase. My suggestion would be to see if you can use STAR is this manner with however you create your workflows.