Using varVAMP on bacteria

[mxwang66]

Thanks for developing varVAMP! In the note it says varVAMP is "specifically developed for viruses", but I wonder if I could also use it on bacterial genomes since they have less variations and should be easier to design. Of course for bacteria you cannot make a whole genome MSA, so I selected a target region (~400bp) and made an MSA only for that region (e.g., a region that is conserved in most genomes in my target species).

I ran varVAMP qPCR on my MSA using -t 0.9 (most MSA cols are >0.98), and it chose -a as 0 (as expected due to fewer variations). At first it couldn't find a design. Then I modified the config file and relaxed primer sizes, GC content, etc., and it did generate one design. I'm posting this issue here to check if there's any precautions to use varVAMP on bacterial genomes, and to help other users with the same question.

p.s., in the log I saw 5 unique amplicons with internal probe, but only one passed deltaG cutoff. I wonder what's the meaning of this cutoff? Is this related to QAMPLICON_DEL_CUTOFF in the config?

Best,
Michael

[jonas-fuchs]

Hi Michael,

thanks for adressing this! You are completely right. The primer design is not constraint to viruses in any way, but it was developed with the variability of many RNA viruses in mind. But you can of course use the tool for more conserved regions in bacteria or other species. For qPCR designs such as in your case, its exactly the way to go. I would only recommend making the alignment a bit larger, that more potential candidates can be found.

The deltaG cutoff is a proxy for secondary structures in the qPCR amplicon. The lower the Gibbs free energy is, the less secondary structures and the better the qPCR amplicon is amplified. varVAMP makes here use of the library seqfold and tests the amplicons deltaG at the lowest melting temperature of the flanking primers. You can increase or decrease this parameter directly in the arguments with --deltaG.

QAMPLICON_DEL_CUTOFF - ok so this is more a cryptic setting which is not so well described in this github. In the alignment-derieved consensus sequence, varVAMP mask gaps with 'N' and is later on aware to not design primers in these regions. Gaps are also relevant for qPCR amplicons. For a qPCR amplicon that spans a region with a really large gap deltaG calculations will not be accurate any more and you might anyways not want a qPCR design that can be larger for some sequences. However, smaller gaps might not be an issue and deltaG calculations are only off by a bit for sequences that e.g. have 3 nts more. The cutoff in the standard settings means that gaps larger 4 nt are not tolerated. varVAMP knows this as these gaps are masked with 'NN' in the conensus sequence: Code reference 1 and code reference 2 if you are interested. Hope this helps ;).

Cheers,
Jonas

[mxwang66]

Thanks Jonas, that's very helpful! I'm actually going to test varVAMP's qPCR design on Campylobacter jejuni, and will let you know how it goes! I'll also try longer regions as you suggested.