Where is the script for Microwell-seq data?

[JianingYao]

Hi,

Thank you for maintaining such a helpful repository! I’m working with Microwell‑Seq data and would like to use STARsolo to process it. Could you please point me to the relevant script(s) in the scripts directory that demonstrate this workflow?

Thank you in advance,
Jianing

[apredeus]

Hi Jianing,

Terribly sorry - I was under impression that I've committed the script for Microwell-seq, but apparently I did not, and now it is lost. I will re-create it in due time; meanwhile, here are the options that I've used for the processing (copied from the logs), and the barcode files in attachment. This was run successfully on a huge atlas done with Microwell-seq (GSE134355). And yes it's not a typo re. using microwell_bcA three times in a row - that's what has been used in this particular dataset.

STAR --runThreadN 16 --genomeDir /nfs/cellgeni/STAR/human/2020A/index 
--readFilesIn <R2> <R1>  --runDirPerm All_RWX --readFilesCommand zcat --outSAMtype None 
--soloType CB_UMI_Complex --soloCBmatchWLtype 1MM 
--soloCBwhitelist microwell_bcA.txt microwell_bcA.txt microwell_bcA.txt 
--soloCBposition 0_0_0_5 0_21_0_26 0_42_0_47 --soloUMIposition 0_48_0_54 --soloFeatures Gene GeneFull Velocyto 
--soloOutFileNames output/ features.tsv barcodes.tsv matrix.mtx --outReadsUnmapped Fastx

microwell_bcA.txt
microwell_bcB.txt
microwell_bcC.txt

[JianingYao]

Thank you so much! This is extremely helpful! I have some follow-up questions:

1. Is there a particular reason to use microwell_bcA three times? When should I use the other barcode files?
2. Also, is --soloCBposition or --soloUMIposition dataset-specific or a general rule for Microwell-seq data?
3. Does R2 or R1 contain cDNA? In the STAR manual, it said for "--soloCBposition: Presently only works with –soloType CB UMI Complex, and barcodes are assumed to be on Read2."
4. I also noticed that microwell_bcB.txt and microwell_bcC.txt are same.

[apredeus]

Hi Jianing,

1. You can just check which whitelist matches each position in the read using Unix tools - and that's what I did. I guess every time someone runs it they can use whatever whitelist they see fit. For this particular experiment, it was the same whitelist 3 times..
2. I did not process enough such datasets to answer this question, sorry :)
3. That's interesting. I think "read 2" in the manual refers not to actual Illumina R2, but to the read that comes second in the STAR command. I've ran it using the command above and obtained good mapping. R1 in my notation is one with BC+UMI and R2 is cDNA read.
4. I guess that makes it easier to run, good spot!