cell barcode question

Question

cell barcode question

Closed this issue 5 years ago · 7 comments

Hi,

Recently I tried to analyze scRNA-seq data produced by mcSCRB-seq. Because I have valid cell barcode file, so I did not need to identify the cell barcode in step2. Is any possible your tool could use the reference barcode file to extract the reads? I tried use " umi_tools extract" with --whitelist=MY_REF_INDEX, however it still failed.
Could you give me some suggestions?

Thank you!

Answer 1 · 2019-10-24T13:25:15.000Z

Hi @skytguuu,

Could you describe what FASTQ files you have and where the UMI/cellular barcodes are in them?

Answer 2 · 2019-10-25T03:53:51.000Z

Hi,
My fastq files is almost 150bp base. Position 1-6 is as the cellular barcode and positions 9-16 is as the UMI. Followed is my fastq file:
###############
@E00477:517:H3CN7CCX2:4:1101:23084:1432 1:N:0:CTCTCT
NGCATGTGTCTACTAGTTTTTTTGTTTTTTTTTTTTTTTTTTTTTTTTAAGTAAACGCTTCCGGGCCCGCGGGGGACTCAGGTTAGAGCATCGAGGGGGGGGCCGGGAGGCAAGGGGGGGGGGTGGGCGGTGACTCGCGCTGGGGGGGAC
+
#AAFFJJJJJJFJJFJ77FFFJJJJJAJJJJJJJJJJJJJJJJJJJJ--<<7-<J-<-<F--7-7-AJ<77AA-7A-7--7-7-7--<-7<<--<<JFFJF--7A-77AJAF--<7A-7<FJ-7-AJ7-AAJ------))))))))--)-
###############
Thanks for your help!
Best

Answer 3 · 2019-10-25T13:31:31.000Z

Heya-- are these single-end reads then? Or do you have paired reads? Usually SCRB-seq has the first read as you described, and the second read is the biological read. Is that the case with your experiment?

Answer 4 · 2019-10-25T16:03:22.000Z

yes，I got two fastq files.R1 and R2, usually R2 was used to do the mapping. 发自我的iPhone

…

------------------ Original ------------------ From: Rory Kirchner <notifications@github.com> Date: Fri,Oct 25,2019 9:31 PM To: vals/umis <umis@noreply.github.com> Cc: skytguuu <616413051@qq.com>, Mention <mention@noreply.github.com> Subject: Re: [vals/umis] cell barcode question (#59) Heya-- are these single-end reads then? Or do you have paired reads? Usually SCRB-seq has the first read as you described, and the second read is the biological read. Is that the case with your experiment? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Answer 5 · 2019-10-25T19:43:36.000Z

Hi @skytguuu,

I see. The first step will be to stick the UMI and the barcode in your FASTQ name for R2, by running this:

umis fastqtransform examples/SCRB-Seq/transform.json fastq1 fastq2 > your-sample-transformed.fq

this will stick the cellular barcode and UMI in the read name.

Then you want to take your list of good cellular barcodes and do:

umis cb_filter --nedit 1 --bc1 your-barcode-file.txt your-sample-transformed.fq > your-sample-filtered.fq

This will remove all of the barcodes from your-sample-transformed.fq that are not one of the barcodes listed in your-barcode-file.txt, but will fix ones that are one edit distance away from a known barcode and you'll be left with just barcodes that are in your-barcode-file.txt. Does that help you get going?

Answer 6 · 2019-10-26T16:32:47.000Z

Hi,
It worked. Thanks for your help!

Answer 7 · 2019-10-28T15:49:15.000Z

Great! Glad that got you going. Thanks for following up!