454 problem with adapter trimming and MIDs

[OllyBolly]

Good idea -thanks.

[JackieBadger]

We had the same problem coming from outsourcing to a 454 Junior run.
Lucky for you that the MIDs are distinguishable at 7bp. Just trim both MIDs down in your barcode file

[OllyBolly]

Thanks AJ.

Actually the barcode splitter/sorter in Geneious that I've used seems to deal with the missing base anyway, so luckily no major disruption, but if not for that I could have ended up with skewed results. Just something for people to be aware of.

[ajthomas]

I haven't ever come across that, and I'm not sure why it might be happening either. My first thought was that the software might be thinking it is part of the adapter sequence for some reason and trim it off, although I can't think of why it would do that. Nevertheless, I'm not sure you need to figure out the cause in order to fix it. If you got the .sff files from your provider, the data is still there. You may be able to put those missing bases back on by changing the trim position with sfffile. If your provider gave you the whole D directory, you could try reprocessing the data, perhaps with the shotgun pipeline or changing other parameters, and see if you can get the data trimmed in the right position.

[OllyBolly]

We've had some odd trimming results and I thought I'd check whether anybody else has had a similar experience.

I received raw 454 reads from our provider with the A & B adapters removed (presumably via the 454 pipeline). I have MIDs on both the 5' and 3' end of the reads to facilitate multiplexing. I've noticed, however, that reads with MID 16, 18, 19, and 22 on the 3' end of the read have their terminal (i.e. 3') T nucleotide missing. Depending on how you demultiplex, this can mean that MIDs are not correctly read & sorted. The problem seems restricted to the MIDs I've used that begin with T

It seems something to do with trimming the B adapter because MIDs that I have on the 5' end of the read, such as MID 10 that also have a terminal T (i.e. 5'), retain the T.

Anybody else come across this?