Oligos that do NOT bind in mouse genome - (Jul/30/2012 )
I need to design oligos that will be used as PCR and sequencing primers and it is important thet they do not bind anywhere in mouse genome. These should be single oligos, not a pair. I want to get PCR products where one primer is specific for a defined sequence (I already have some good primers for this) and the other has to be non-mouse.
I designed some primers in Roche Design Center (other organism) using some plant specific genes. But when I blast the primers against mouse genome, I always get hits e.g. 16 from 20 nucleotides primer show max identity with some mouse genes. I'm afraid this may be enough to cause unspecific PCR products.
Is there any website or any program, that would generate primers that definitely do not bing anywhere in mouse genome?
Are those genes you want to sequence honmologous with mouse genes? I mean like, if it's some basic metabolic pathway gene or something, it probably will have homologs. Mostly problem with nonspecific binding is within homologous genes or pseudogenes.
You can past those sequences (plant and homologous mouse) to Roche Design Center in fasta format, and select Differentiating assays, that should design primer pairs that don't bind the other sequence. You can BLAST with mouse and the plant these sequences to check for specifity within a species and in mouse.
BLAST will mostly show some minor binding depending on the relative abundance of certain sequence pattern. But most of these doesn't really matter, saying 16 out of 20 bp doesn't actually say much, when you don't specify WHERE in the sequence it binds.
Primers have to be specific on their 3' part, mismatch near/at 3' end, as small as 2 bp can cause complete loss of binding even if the rest is complementary.
If you want to post your BLAST ID or the sequence you got, I could check it. Also stringency of the conditions (annealing mainly) is important for primer specifity, for that reason I usually design my primers for higher Tms like 65, but that's a matter of opinion. Of course always test your primers for specifity (negativity on a pure mouse sample).
Anyway designing specific primers within species is possible. I designed several primers for human/not-mouse even in such highly identical genes as beta-globin. That though required a pretty painfull manual designing, but I don't think that between plant and mouse this would need to go that far. Maybe your first primers are OK, you just got scared by BLAST
The point is that my primer should not bind anywhere in genome. I'm not working with a specific mouse gene, but I will work with random fragments of mouse genome. So I'm not sure that if I do differentiation assay will help me much, I would have to input the whole mouse genome.
But definitively I will use high Tms, that's a good idea.
Ah.. stupid! I completely forgot, the PrimerBLAST tool can design on any given sequence, so like a plant one, by BLASTing to any organism database, like a mouse, and chose only the specific ones!
You can choose genome or transcript database. It uses primer3 algorithm for primer design, but counts Tm differently. If you are accustomed to primer3 calculation (as I am, usually use Ta 4-5 degs lower than Tm given by primer3) you need to select Salt correction formula - two guys 1965 and Table of thermodynamic parameters - Breslauer 1986 in Advanced parameters.
Hope this helped, finally..
Yes, thanks, you helped me a lot