Star solo , cellranger

[saassli21]

Good morning , I'm new in the field of bioinformatics and I want to verify some points with you. what are the 2 approaches adopted by cellranger and star solo ? is it true that CellRanger uses a reference-based approach to align RNA-seq reads whereas StarSolo uses a de novo approach generate the count matrix ?

another question: is there an equivilant to cellranger aggr if we want to generate a count matrix from multiple samples ?

[alexdobin]

Hi @saassli21

STARsolo can be run to produce results very similar to CellRanger. They both use STAR to map reads to the genome and then assign the alignments to annotated genes.

There is no aggr option in STARsolo. You would need to downsample the samples that have too many reads, and then run STARsolo on each sample.

[DaliBAmor]

Sir, thank you very much for responding since I'm new in this field I didn't really understand what do you mean by downsample the samples that have too many reads ? and how to do that ?

• since both cellranger and star solo uses reference based approach and they both use STAR as aligner why star solo is much faster ( I searched the information on google but I didn't find one )

[alexdobin]

Downsampling means selecting randomly a subset of reads from each sample, so that the number of reads in each sample is the same. STARsolo is several times faster than CellRanger because it's algorithm for counting UMIs is more efficient.