Spike-in for identification of samples

[hbn]

Hello everyone,

Does anyone know if it is possible to spike-in your NGS samples with different spike fragments, to reconfirm that your sample is the right sample, based on the spike sequence?

So basically what I'm looking for is a set of different spike sequences that you can add to different samples to check if there hasn't been any mix-up in some step.

Hope anyone can help...

[pmiguel]

Are your samples RNA or DNA?

Are you testing for mix-ups from those doing library construction, or just loading into the wrong lane of the flow cell?

In principle it should be possible. Could be issues, however if, for example, you spiked-in too much and most of your sequence came back spike-in. Also, RNA would be more problematic if you are not making the libraries yourself because your spike-in would probably need to be polyadenylated so that it would not get lost during ribo-depletion.

All in all, I am not a big fan of spike-in's during library construction because in cases where the quality of the incoming samples are marginal, spike-ins can end up dominating the library. And down the line some poor schmuck is going to end up thinking your spike-ins really derive from your sample and draw all kinds of incorrect inferences. Especially problematic if the poor schmuck is you a couple of years later...

--
Phillip

[hbn]

Basically, I'm testing for mix-ups for different steps in the procedure. If you get sequence results from sample 1, that you can see from the spike that it is really sample 1.

Are there any commercial kits available for this? (I haven't found any so far)

Samples are DNA or cDNA, but I guess that should not be much of a problem...

I can imagine that it is very crucial to use a very low concentration, but I guess that can be calculated.

[pmiguel]

I am unaware of any sample tracking spike-in reagents. Illumina TruSeq kits now include spike-in reagents that provide diagnostics on success/failure of various steps in library construction. Possibly they could be modified for sample tracking.

--
Phillip

[HESmith]

An alternative would be to barcode your samples as you would for multiplexing.