search

alexdobin / STAR

RNA-seq aligner
MIT License
1.78k stars 497 forks source link

No mapped reads 10x genomics 5' single-cell RNAsequencing #1868

Open nicholasschwab opened 1 year ago

nicholasschwab commented 1 year ago

Hello everyone,

I cannot figure out where I go wrong, but I get no mapped reads from 10x fastq files that I downloaded from SRA.

example:

https://www.ncbi.nlm.nih.gov/sra/SRX8976950

I used this STAR command, which has worked in the past for 10x 5', but maybe there is a problem with the chemistry version that I do not see?

Thanks for any help!

STAR --soloBarcodeMate 1 --clip5pNbases 39 0 --soloStrand Forward --soloType CB_UMI_Simple --soloCBwhitelist /STAR/counts/737K-august-2016.txt --soloCBstart 1 --soloCBlen 16 --soloUMIstart 17 --soloUMIlen 12 --soloUMIdedup 1MM_CR --soloCBmatchWLtype 1MM_multi_Nbase_pseudocounts --soloUMIfiltering MultiGeneUMI_CR --soloCellFilter EmptyDrops_CR --outFilterScoreMin 30 --soloFeatures Gene GeneFull Velocyto --soloOutFileNames test/ genes.tsv barcodes.tsv matrix.mtx --genomeDir /STAR/counts/STARgenome --runThreadN 56 --readFilesIn fastq_R1 fastq_R2 --soloMultiMappers EM --outReadsUnmapped Fastx

alexdobin commented 1 year ago

Hi @nicholasschwab

Which protocol is this sample and which chemistry?

nicholasschwab commented 1 year ago

It is supposed to be a 10x genomics 5' (with BCR sequences in a separate fastq file) and the authors do not specify which chemistry. However, I have previously successfully converted V3 chemistry with STAR solo and, therefore, I don't think it's V3. I have tried to adjust for V2 with other UMI length, but to no avail.

A read looks like this (if that helps):

gnl|SRA|SRR12483421.1.3A00351:91:HGNMFDSXX:1:1101:10004:10207 Biological (Biological) GTTGACAAAGAAAAACAAGGCATCCTCAGCTCGGAGATGAATTCGCTTCCGGATCAAGAAGTAGAACTGACCAACTGTGAGATCAGAAGGCACCAGGTATTTCTTTTTGTCCAGGTCTCCTATCCGAGCTTTGGGAGCCTTTTCTACTAT

alexdobin commented 1 year ago

Since they have CB/UMI read separate from cDNA reads, I would try to treat them as if they were 3' v2, i.e.

--readFilesIn read_150b read_26b
--soloCBstart 1 --soloCBlen 16 --soloUMIstart 17 --soloUMIlen 10

without --soloBarcodeMate 1 You may need to try both --soloStrand Forward and --soloStrand Reverse to get the right strandedness. Also, you may need to try different passlists.