search

alexdobin / STAR

RNA-seq aligner
MIT License
1.82k stars 502 forks source link

How can I use starsolo for single cell combinatorial indexing RNA sequencing? #2028

Open caodudu opened 8 months ago

caodudu commented 8 months ago

I noticed starsolo was designed to work in many different platforms of scRNAseq, e.g. 10X/Indrop/SeqWell. However, I don't find the corresponding category for single cell combinatorial indexing RNA sequencing including sci-rna-seq/SPLiT-seq/sci-rnaseq3. Their cell barcodes are combined by separated parts in two or more reads(R1/R2/I1/I2).

  1. So if I want to implement starsolo for them, What should I do?
  2. I heard Cellranger claimed the base quality of truncated read1 doesn't pass , and Cellranger would just choose read2 to map , if the input data is from 10x 3' platform. Will starsolo process the reads from 10x 3' like Cellranger or not?
  3. Can starsolo accept modified fastq whose headers are stitched with CB and UMI by other scriptts？
alexdobin commented 8 months ago

STARsolo can work with simple and complex CB/UMI geometries. Please check the documentation: https://github.com/alexdobin/STAR/blob/master/docs/STARsolo.md If you can transform CB and UMI to one of the acceptable shapes, it should work.

caodudu commented 8 months ago

STARsolo can work with simple and complex CB/UMI geometries. Please check the documentation: https://github.com/alexdobin/STAR/blob/master/docs/STARsolo.md If you can transform CB and UMI to one of the acceptable shapes, it should work.

Thank you. I will try. But the full length of CB from sci-RNAseq3 is 40. Is it longer than the limitation(32-mer) of starsolo?

ONeillMB1 commented 7 months ago

@alexdobin thank you so much for developing this software. I'm wondering if you have any plans to more formally accommodate combinatorial indexing methods. From discussions in various issues, I can see that there are solutions provided for Parse/split-seq, but not Scale/sci-RNA-seq.

alexdobin commented 7 months ago

Hi @ONeillMB1

At the moment I do not have bandwidth to do it. These companies have their own pre-processing scripts that are tuned specifically for their most recent protocols, and it's best to use them.

jpcleland commented 4 months ago

@caodudu My colleague Job van Riet (with input from Lauren Saunders) has put together a sci-RNA-seq3 processing pipeline which, among other things, uses STARsolo for alignment and UMI counting. It's not formally published yet, but code and documentation has been up on github for a while (https://odomlab2.github.io/sci-rocket/) and there is some additional detail in the M&M of a recent preprint: https://www.biorxiv.org/content/10.1101/2023.12.04.569933v2. Please give it a shot and let us know what you think!