Different barcodes on demultiplexed data

Question

[pablo]

Hello,

I ran a demultiplexing set on 12 samples with the SMRTtools (PacBio). For that, I used 12 different barcodes in FASTA format.

All of the 12 barcodes are found, but there is an extra barcode found (different pair in the code below). I guess it disturbs the demultiplexing analysis.

With same pair                : 647416 (100%)
With different pair           : 1992 (0%)
Coefficient of correlation    : 136.93%

Why there is an extra barcode found, while in my other demultiplexing analysis, only the barcodes used during the library prep are found?

Bests

[benformatics]

How similar is the extra barcode to your other barcodes? Could it be due to sequencing error?

[pablo]

There are two different pairs :

>bc1011
CTATACGTATATCTAT
>bc1126
GTGTATCAGCGAGTAT

and

>bc1012
ACACTAGATCGCGTGT
>bc1126
GTGTATCAGCGAGTAT

They look similar.

I also noticed there is one pair bc1126--bc1126 which matches with around 1000 .bam sequences.

Could be a contamination with the bc1126 barcode during the library prep?

[pablo]

There are two different pairs :

>bc1011
CTATACGTATATCTAT
>bc1126
GTGTATCAGCGAGTAT

and

>bc1012
ACACTAGATCGCGTGT
>bc1126
GTGTATCAGCGAGTAT

They look similar. Could be a contamination with the bc1126 barcode during the library prep?

[GenoMax]

Could be a contamination with the bc1126 barcode during the library prep?

Possible. I have not used PacBio barcodes but having a small smattering of unexpected indexes is normal in Illumina sequencing. Perhaps same applies here. Number of reads looks small enough to ignore. You did use demux tools provided by PacBio? We had discussed that in a past thread.

[pablo]

I suppose it is a contamination yes. I used a tool provided by PacBio right.

Could you suggest me a demultiplexing tool not provided by PacBio, to try it?

[GenoMax]

I suggest you don't bother with trying to find other tools. If you have demultiplexed your data go on with other analysis which would be far more interesting.

[pablo]

Actually, I did a second analysis by removing the "extra barcode" . It also worked well. I will keep that analysis.

Thanks for your help.