Paired ends demultiplexing Qiime Illumina

Question

[carvin900]

Hi,

Sorry in advance for my beginner level in bioinformatic. I need to do analysis of illumina reads paired end. The starting files I have are: fwd.fastq , reverse.fastq and mapping file.txt.

In the sequences files, there are several samples with different barcodes (barcodes with different lengths) and different primers (3 different set of primers). I tried so far to:

1. I extracted barcodes with extract_barcodes.py --bc1_len X --bc2_len X (where is x is the length of the barcode). I have barcodes of length 4,5,6,7 and 8 so I did 5 times this command with different lengths
2. I merged the reads1.fastq and reads2.fastq output files from 1) for each barcodes length.I had 1,78% in general of non-merging sequences.I also asked to get a fastjoin.join_barcodes.fastq file.
3. I tried to do split_libraries_fastq.py with the merged sequences and the fastjoin.join_barcodes.fastq file. But then I get over 1 900 000 reads non associated to barcodes out of 2 500 000 total reads.

If someone could help me I would be really glad!

Vincent

Answer 1

[GouthamAtla]

Why don't you use illumina cassava for demultiplexing the data ? Is it a GBS kind of data where you have internal barcodes other than standard barcodes used for multiplexing ?

[carvin900]

Hi Geek_y,

thanks for the rapid answer. I am fairly new in bioinformatic and my research lab is mostly using Qiime, so haven't heard of CASAVA. Also, our local server is based on Qiime. Would it be possible to launch commands from Qiime to CASAVA? Thanks for your help!

Vincent

[carvin900]

Thanks for your help Geek_y, I found out my mistake. There was a switch of two bases in our primer sequence registered and the primer sequence in the reads. Now all is doing fine. I only have done merging and already split_libraries.py. Thanks again!