RAD libs - bands in the test amplification step

[mariesan]

Hello all,
has anybody experienced distinct bands when making RAD libraries? They appear in the test amplification step. I get a smear between 200-700 bp and on top of that the bands. I've made a number of libraries, but this is the first time I've seen this. Contamination? Hope not, because I really wouldn't know where to start...

[nucacidhunter]

You probably should not sequence those libraries. Distinct bands are indicators of repeat region presence (transposons, organelle DNA). Are these classic RADseq or other variants of the technique?

[SNPsaurus]

What enzyme(s) are you using? As nucacidhunter says, probably not a good sign. Just as a guess, if this was a SbfI RAD-Seq (as opposed to some variant) library, you could see something like that if a repeat had a 1.5 kb SbfI - SbfI fragment. It would shear poorly because of the small size, and then amplify as a distinct band because of the many copies present in the genome. It could mean the smear is OK, but could also suggest that some aspect of the library making is inefficient, letting these artifacts become more dominant.

[mariesan]

Yes, classic RADseq. I use the protocols from Etter et al and Amores et al 2011, with some modifications. Cutting with SbfI. The samples has been sequenced and turned out poorly, according to the sequencing lab, because of "many short fragments". So I did them again and this time the bands appeared clearly. Makes sense what you both suggests.
I'm just surprised that the problem turned up all of a sudden. I treat these samples the same way (using the same reagents) I do with all other samples of this organism. So why just these two samples out of many? I know, impossible to answer : ).

[SNPsaurus]

What species are you genotyping? Sometimes a few samples will be contaminated with something else (for example, an insect will be infected with a fungus). You could grab some of the sequences from the previous attempt and blast them to see if they match your species or something else. Also, did you look at the DNA from those samples on a gel or Fragment Analyzer and notice any difference between them and the other samples that did work well? But, probably hard to definitively find a cause.

[nucacidhunter]

I wonder if you could post an image of your library. With RADseq it is less likely to see bands as a result of repeat region. I agree with SNPsaurus explanation.

[mariesan]

I'm working with phytoplankton. The cultures are clean, apart from some bacteria maybe. Maybe it's enough for showing up as bands? We will have a closer look at the sequences as suggested.
Unfortunately, I only have a lousy gel photo that doesn't help much. There's a smear in the background and bands at 450 and 350 bp.
Many thanks for your answers and time spent.