should primers be considered biological sequence?

[sformel]

Hi,

Forgive me if this is naive, but I'm still newish to thinking about sequencing. I've begun working with the DADA2 software and the tutorial clearly states that primers should be trimmed because they are non-biological nucleotides.

I'm not arguing against it, but I don't fully understand it. If the primer matches a sequence of DNA, then shouldn't it be just as legitimate a sequence as the sequence derived from the sequencing process?

Any thoughts?

Thanks,
Steve

[jhalpin]

Primers and adapters are added to the fragments during the preparation of the samples. They are not native sequences and are extremely over represented compared to any natural occurrences of those short sequences.

[sformel]

Thanks for the quick reply! But I'm still not sure I understand.

My thought is that the because the primer matched the DNA fragment during PCR, it can be considered a legitimate copy of a biological sequence of nucleotides, just like the copy that is made during sequencing.

But it seems from what you wrote that the problem is due to over representation of those sequences. How is a primer sequence different from a highly conserved region of an amplicon?

To be clear, I'm coming at this from a metagenomics perspective, which may be skewing my thoughts.

[jhalpin]

Sorry.. in the lab to make sequencing libraries the primers are added. They aren't in the sequence of the thing you're looking at. The primers are constructed by a company and sold as part of the kit. Here's an overview of the lab side: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4351865/

[sformel]

Ok, I understand what you're saying in regards to RNA-seq type libraries.

However, I'm working with a 2-step PCR in which the first step is an amplification of a marker gene, followed by an indexing PCR (addition of adapters and barcodes). So in my case, the primer is part of my biological sequence.

In any case, I appreciate your thoughts. I'm going to think on it a little more and see if it clicks.

[cmbetts]

That sequence isn't derived from the fragment you're trying to sequence. 100% of it was chemically synthesized by your oligo synthesis company. It match close enough to your insert of interest to prime it, but does not guarantee that the sequences were perfectly complementary. For instance, any mispriming artifacts will perfectly match your sequence of interest over the length of the primer...

[sformel]

Ah ok, that makes sense. I didn't think about those kind of biases. Thanks for your help!

[ATϟGC]

I will just echo cmbetts in saying that you should always remove your oligo/primer sequences from Sanger and High-throughput sequencing data since they are artificial. You can sometimes find primer/oligo sequences in "cleaned" sequences archived in GenBank and other public databases but this is not a good practice.

The only good reason I can think of off-hand to not trim the oligos is if you are archiving the raw data, in which case you should state the oligos used so others can trim them for downstream analyses.