Filtering reads from aligned scRNA reads BAM file by filtered cells barcodes (CB)

Question

[MYousry]

Hello Everyone,

I passed my scRNA fastq data (R1 & R2) through STARsolo to align the reads with genome and produce the count matrix. I want to filter the resulted BAM (Aligned.sortedByCoord.out.bam) using the filtered cells barcode (solo.out/gene/filtered/barcodes.tsv) so as the resulted BAM file contains data from actual cells.

I've tried 10x genomics' subset-bam tool but everytime I run it, I get the following error:

Any help with that would be greatly appreciated! A beginner here:)

Best, Yousry

[Pierre Lindenbaum]

save the planet. Why a screenshot when you can just copy-n-paste the text ?

[MYousry]

subset-bam --bam Aligned.sortedByCoord.out.bam --cell-barcodes barcodes.tsv --cores 1 --out-bam /results/ --log-level debug

[rpolicastro]

STARsolo has a 10X-esque barcode correction and cell filtering algorithm built in. See their documentation for more information, specifically the --soloCellFilter and --soloCBwhitelist arguments.

[MYousry]

Thank you so much for your reply and guidance! I will look into that.

[MYousry]

I have tried it but seems like the outputted BAM file is the same? I think these commands has to do with the count matrix only, no? I am interested in getting a filtered BAM like the one produced by subset-bam. Can you please guide me through the process?

these are the commands I ran: STAR --runThreadN 4 --genomeDir Reference/ --readFilesIn FASTQ_data/SRR13040580_2.fastq FASTQ_data/SRR13040580_1.fastq --outFileNamePrefix STAR_SoloCB/ --outReadsUnmapped Fastx --outSAMattributes NH HI NM MD CB UB sM sS sQ --outFilterMultimapNmax 1 --outFilterMatchNmin 30 --outFilterMismatchNmax 4 --alignIntronMax 1 --alignSJDBoverhangMin 999 --soloType CB_UMI_Simple --soloCBwhitelist Whitelist/V2/737K-august-2016.txt --outSAMtype BAM SortedByCoordinate --soloCellFilter EmptyDrops_CR

[GenoMax]

Looks like your command line options are not correct. Have you looked at in-line help for correct options.

Keep in mind that 10x tool you are working with may only be designed to be used with cellranger generated BAM files.

[MYousry]

Thank you for your reply!

I tried to check for errors in the command line but I couldn't find any. I think differences in the generated BAM files might be the issue. Any idea how I can get this to work (e.g any other tool or edits to this command line)? Thank you!

[MYousry]

So, yes, i had an error in the command line (clarified it in the comment below). Thank you!!

Answer 1

[MYousry]

So it turned out that I have an error in the command line I wrote. --out-bam should take an argument for the name of the outputted bam file not the directory of the output. I removed the "/" and it worked!

Thanks for the support!

myousry@MYousry:~/BAM_plz$ subset-bam --bam Aligned.sortedByCoord.out.bam --cell-barcodes barcodes.tsv --bam-tag CB:Z --log-level debug --out-bam filtered

13:07:55 [DEBUG] subset_bam: Loaded 2150 barcodes

13:12:08 [INFO] Chunk 0 is done

13:12:09 [INFO] Done!

13:12:09 [INFO] Visited 14050588 alignments, found 14050588 with barcodes and kept 11545307