Illumina paired-end sra data in three separate files - what next?

[JonB]

Hi,

I have used fastq-dump to split paired-end illumina data. I get three files, one for each different pair and one file with barcodes. This is transcriptome data and I want to do de novo assembly. I have two questions:

First, on the SRA website where I got the data it is only mentioned one barcode while there are several different in the barcodes file. Should I only use the sequences with the barcode given on the web?

Second, how can I split the files according to the different barcodes while keeping the pairs? I looked at the fastx toolkit and the qiime split_libraries, but I don't think my illumina barcodes are inlcuded in the sequences themselves?

Examples of the files:

-bash-4.1$ head SRR343051_1.fastq 
@SRR343051.1.1 B0A05ABXX110604:3:1101:18610:1087 length=101
NTCTTCTTGCGTACGCATTTGGACTTAATCCTAATCTTGGATTTGTTTCTTCTAAATATGTACCAATCACAATGCTTGAATCTCTTATTATAATATATTTA
+SRR343051.1.1 B0A05ABXX110604:3:1101:18610:1087 length=101
#####################################################################################################
@SRR343051.2.1 B0A05ABXX110604:3:1101:14471:1088 length=101
NCGAAGGGCAATGTAATAAAGTTTATTATTATGTGTGTACAATGCAAAAAAAAGGGACTCGACTCTAATCCTGGTCGAAGCACAGGGCAAGACCACCAATG
+SRR343051.2.1 B0A05ABXX110604:3:1101:14471:1088 length=101
#####################################################################################################
@SRR343051.3.1 B0A05ABXX110604:3:1101:20187:1088 length=101
NATCATAATCTTCAATTTTCAAATTACTCTTGTTGCCTTTGGAAAGATCGTTAGTTTTCGGGTCTTTTATATTTTACTATTGCTTTATACTTGTTTTCACT

-bash-4.1$ head SRR343051_2.fastq 
@SRR343051.1.2 B0A05ABXX110604:3:1101:18610:1087 length=8
TTGAGCCT
+SRR343051.1.2 B0A05ABXX110604:3:1101:18610:1087 length=8
CCCFFFFF
@SRR343051.2.2 B0A05ABXX110604:3:1101:14471:1088 length=8
TTGAGCCT
+SRR343051.2.2 B0A05ABXX110604:3:1101:14471:1088 length=8
CCCFFFFF
@SRR343051.3.2 B0A05ABXX110604:3:1101:20187:1088 length=8
TTGAGCCT

-bash-4.1$ head SRR343051_3.fastq 
@SRR343051.1.3 B0A05ABXX110604:3:1101:18610:1087 length=101
GAGAAAATAAAATATGAGAAAATAGTAAAGAAGAAATTAACTGATATAATTACAGAAGAGAATGAATAATTGAAACAATTAAAAAATCATTAAATGAAGAT
+SRR343051.1.3 B0A05ABXX110604:3:1101:18610:1087 length=101
CCCFFFFFGHHHHJJJIJIJJIJJJHJIJJJJJJJJJJJJJJJJJJJJHIGIIIIGHHIJIJJJJJJIJJJJJEGIIJJJJGFHHFFCEEEECCDDDCCCC
@SRR343051.2.3 B0A05ABXX110604:3:1101:14471:1088 length=101
CTGATGGTGTACGTTGAACTTGGTCTGGTGGTGCTGATTCTGAGCAACAGTCTGCGTCGCGCCGCCTCCTTCTTCCTGATTCTCTCGCTGGCCGTGTCGCT
+SRR343051.2.3 B0A05ABXX110604:3:1101:14471:1088 length=101
BCCFFFFDHHHHHJJIIGIJJJJHIJJIIJJFHIJJIJJJJIIJJJJJJJJIIJJIGIJJHFFDDDBDDDDDDDDDDDCDDDDCDD<BD39??&09B?9A<
@SRR343051.3.3 B0A05ABXX110604:3:1101:20187:1088 length=101
AGGTGATTCATCATCTTCAAAATATTAATAAAAAGTATATTAATATAAAGACAATTATATATCGAAAGTGAATAGTACTGTGAAGGAAAGTAGGAAATATT

[GenoMax]

Hopefully you have the information about barcode <--> sample.

Try this script for demultiplexing: http://qiime.org/scripts/split_libraries_fastq.html

[GenoMax]

@Jon B: You have not used the

-F | --origfmt Defline contains only original sequence name.

option with fastq-dump so you have the SRR* in the names. Just keep that in mind.

[JonB]

@Jon B: You have not used the



option with fastq-dump so you have the SRR* in the names. Just keep that in mind.

Thanks! I didn't see that option.

[JonB]

Hopefully you have the information about barcode <--> sample.

Try this script for demultiplexing: http://qiime.org/scripts/split_libraries_fastq.html

GenoMax, do you mind telling me how I could use this script? I was looking at it before, but I don't understand how it assigns my reads into files based on the barcodes, and how does it deal with the two read pairs? Can I still use it on my data with the pairs in separate files?

Thanks

[GenoMax]

Jon: This appears to be a single sample even though the barcode read is included as a separate file in the SRA archive. See the corresponding ENA record (http://www.ebi.ac.uk/ena/data/view/SRR343051).

In short, demultiplexing is not needed for this sample. You can use the _1 and _3 files as the R1/R2 read pair.

[JonB]

Jon: This appears to be a single sample even though the barcode read is included as a separate file in the SRA archive. See the corresponding ENA record (http://www.ebi.ac.uk/ena/data/view/SRR343051).

In short, demultiplexing is not needed for this sample. You can use the _1 and _3 files as the R1/R2 read pair.

Thank you!