Barcode Contamination - iChIP

[jeremy.lopez]

Hi everyone,

I have attempted to perform indexing-first ChIP (iChIP) as described here: http://www.ncbi.nlm.nih.gov/pubmed/25103404. The way this protocol works is you perform a total H3 IP, then you index the DNA using the illumina adaptor and unique barcodes, then you pool the indexed chromatin and perform an H3K27Ac IP on the pooled chromatin.

After submitting for sequencing, my barcodes were found in the sample, but additional Illumina barcodes were also found. I never even purchased the barcodes that were found so I was unsure how they might have ended up in my sample?

Does this seem like a contamination at my sequencing facility, or is there any other way for other barcodes to end up in your sample?

[esherman]

Which instrument was used to sequence your samples? The MiSeq, NextSeq (I think), and HiSeq (on rapid-run mode) all perform on-board cluster generation, which means that each library is pumped through the instrument fluidics system. It's not uncommon to see a trace amount of DNA fragments from the previous library leach into the fluidics tubing during cluster generation and leach out of the tubing during the next run. This results in run-to-run crossover. Illumina has released protocols that allow the user to 'bleach' the fluidics system after each run, and this has resulted in a drop in our observable run-to-run crossover.

If your sequencing center used a cBot to cluster your sample on a flow cell, then the sample leaching in/out of tubing hypothesis is invalidated. Thus, the errant barcodes likely originated from the Illumina production line or from the sequencing center during library quantification.

Do the sequences associated with the errant barcodes 'make sense' for your sample? If they map to a different organism or otherwise don't match the expected results from your experiment, then the contamination is likely originating at the sequencing facility.