Drop-seq cell barcode whitelist

Question

[Kevin Blighe]

G'morning,

I have been searching for a cell barcode (CB) whitelist for Drop-seq, but it seems difficult to find a standard list. On the other hand, it is relatively easy to find whitelists for 10x protocols.

If no whitelist is available, what would be the best procedure here? I have used STARsolo to align the reads so far, and it accepts a parameter configuration of --soloCBwhitelist None if no whitelist is available. For all intents and purposes, I presume that I can then just assume that the cell barcodes that STARsolo identifies from one sample is a whitelist that I can use across all samples?

I presume that other programs, e.g., UMI-tools and Alevin, can also take a FASTQ pair and be used here?

Happy Easter / 复活节快乐

Kevin

[ATpoint]

Not really answering your question but as an alternative you could use Alevin for the lowlevel processing. It has a --dropseq flag and (from what I understand) uses a ML approach to decide which CBs are present and trustworthy, so no need for an external list. That is then probably similar to what UMI-tools does since Alevin uses that internally from what I know. Rob and i.sudbery will know for sure.

[GenoMax]

Apparently kallisto may be of help: https://github.com/BUStools/bustools/issues/10 or umi-tools too: https://github.com/CGATOxford/UMI-tools/issues/97

[Kevin Blighe]

Thank you, genomax

[tofukaj]

Dear Kevin,

Would you kindly tell me if you successfully apply STARsolo with Drop-seq? I also plan to learn how to use STARsolo with Drop-seq analysis. It would be very kind of you if you can provide me an example for such application.

Best regards, Kaj

[Kevin Blighe]

Hey Kaj, if I recall correctly, I eventually used --soloCBwhitelist None, which instructs STAR to find cell barcodes automatically. There was a thread on GitHub: https://github.com/alexdobin/STAR/issues/584

[tofukaj]

Thank you very much!!

[tofukaj]

Dear Kevin,

Would you please enlighten me a little bit more? After STARsolo alignment, how could you acquire the expression matrix of all cells of a file? I have tried --quantMode GeneCounts, however it rendered me the overall gene counts of the sam file not gene count per cell.

Best regards Kaj

[Kevin Blighe]

Hey Kaj, after STARsolo alignment, we then use featureCounts

[tofukaj]

Thank you very much!!

Answer 1

[swbarnes2]

Drop Seq doesn't have a white list of barcodes. I think the process by which the barcodes are made doesn't restrict them to a white list.

I use umi_tools whitelist, and umi_tools will also help you pick where the cutoff is between good cell barcodes and empty ones.

[Kevin Blighe]

Thanks swbarnes2