Hi. I have paired end reads (75 bp) from a Nextseq run. Read 1 is 18 bp and Read 2 is 52 bp. R2 is where all the cDNA information is, and R1 is UMI (8) and Barcode (10bp). I've already run umi_tools to detect the UMIs and append those to read file info (https://github.com/CGATOxford/UMI-tools). So R1 now just has the 10 bp barcode left. I think the most efficient way is to take that 10 bp barcode and join it directly to the R2 read on the 5' end. Then I can take the joined R2+R1 fastqs through a program likereaper to demultiplex based on barcodes.
So my question is, what tool can I use for a simple join like that? I've seen tools like ea-utils that seem to do a join and merge based on overlapping bps between R1 and R2, but that's not what I want. I just want to add R1 to R2, no merge or overlap. Any guidance would be great.
1 answer
Moving your answer here (with my modifications).
join <(zcat AC1-10_S34_R1_001.fastq.gz | nl ) <(zcat AC1-10_S34_R2_001.fastq.gz|nl) | awk -F ' ' '{if ($2 ~ /^@NB501800/ ) {print $4" "$5;} else {print $2$3;}}' | gzip > new.fastq.gz
Here is the output:
@NB501800:5:HKYVYAFXX:1:11111:19375:5952 2:N:0:TCCGGCTTAT+GGACTCATTG
GTGCGGATGATCTTACGCTTGTAGGCCAGCCTGGGTGGATATATATTGTGTTCCAAGCCAACTTGGTCTA
++
AAAAAEEEEEEEEEEEEEAAAAAE/EEE/<EEE/<E/EEEAEEEEEA/EE<EEEEE/EEEEEEA/AEEEE
Log in to answer this question.
-
Extract the true single-cell RNA sequencing reads for running SAHMI
written by 573704669 •
-
UMI extraction from 10X visium spatial transcriptome data
written by Omics data mining
-
Queries on 10x sequencing 3' Read 1
written by Cheng Wei •
-
demultiplex paired end fastq file with mixed samples with barcode
written by hellocita •
-
umi_tools using regex greedy quantifiers only returning 1 character
written by williamtmills •
-
Demultiplexing 10x fastq files by GROUPS of barcodes
written by bds1217 •
-
Demultiplex Illumina run using custom index configuration
written by alba.rodriguezmeira •
-
Demultiplexing and deduplicating using barcodes and UMIs in the "mate" read
written by lech.kaczmarczyk •
-
Trim read-specific adapter in paired-end reads
written by Nicolas Rosewick
-
cutadapt --cut option for paired-end reads?
written by blur
Which of the two read headers would you want to keep?
Hi. Thanks for the reply. R2 header.
I am sure there is a
pastesolution that will work. @Pierre, the master of these one liners should be by soon.Here is a bad solution until someone improves on it (you will have to uncompress the files, replace
YOUR_SEQUENCER_ID)Thanks. I'll give it a try. But can I use .fastq.gz? Sorry, I'm not a Unix pro, so this is my attempt:
join <(nl gzip -cd |cat AC1-10_S34_R1_001.fastq.gz) <(nl gzip -cd |cat AC1-10_S34_R2_001.fastq.gz) | awk -F ' ' '{if (@machineID ) {print $4" "$5;} else {print $2$3;}}' > new.fastqSorry I meant to say
sequencer ID(not your machine ID in the sense of computer). It should be the@STRINGat beginning of all read headers. Changed.It worked! Very nice. I need to spend more time learning the power of
awk. Here is my update for for.gzinput and output. I will probably now make some kind of loop to process all my files. Thank you!