Primer BLAST and NetPrimer vs Primer3Plus - (Jul/06/2011 )
Dear all,
when I design primers for quantitative PCR using the Primer3Plus software and then run them via BLAST I get message that they are not specific.
Then I tried the Primer BLAST of NCBI to pick specific primers, and checked them in NetPrimer for seconadry structures etc. I chose the highly rated pairs, which are supposed to have higher amplification efficiency.
Then I run these pairs in Primer3Plus (just to see what will happen) and I get a message that primers are unacceptable because of ustability, complementarity and so on.
1. What is the reason for this contradiction between both programs?
2. Which one is more reliable?
3. The free energy of the hairpins, dimers and crossdimers - the lower it is, the more stable the structure is, is that correct? For example, haipin with deltaG = -11.9 kcal/mol is more stable than hairpin with deltaG = -3.16 kcal/mol, or?
I don't use NetPrimer and I always forget the deltaG values that are suitable, nevertheless.. I design primers in the original primer3, but that has same default settings as Primer3plus, I just like the text output better.
I put my designed primer3 primers to PrimerBLAST to see if there are some SERIOUSLY matched sites, it will always match something, but important is the length (more than 2k will not amplify if your original product is small and you use the cycler settings for small amplicon), the 3' mismatch - if you have mismatch on terminal nucleotide it won't amplify, and the annealing temperature (I design primers to have Ta around 60, which usually gets a decent specificity) of the primers, (and of course setting the right database for specifity RNA/genomic) just note that PrimerBLAST will show a different melting temperature. PrimerBLAST uses the same algorithm as primer3, but has different default settings (which you can change, for example the Tm calculation method), but since it has a limited options I don't use it as a primary primer design tool, unless I want intron spanning real-time primers, which are difficult to design in primer3.
Can you post a pair of your primers from Primer3Plus you're not sure about (and the amplicon sequence), I can check if they're specific on BLAST with my very humble somewhat experienced opinion.
Trof on Wed Jul 6 20:48:06 2011 said:
I don't use NetPrimer and I always forget the deltaG values that are suitable, nevertheless.. I design primers in the original primer3, but that has same default settings as Primer3plus, I just like the text output better.
I put my designed primer3 primers to PrimerBLAST to see if there are some SERIOUSLY matched sites, it will always match something, but important is the length (more than 2k will not amplify if your original product is small and you use the cycler settings for small amplicon), the 3' mismatch - if you have mismatch on terminal nucleotide it won't amplify, and the annealing temperature (I design primers to have Ta around 60, which usually gets a decent specificity) of the primers, (and of course setting the right database for specifity RNA/genomic) just note that PrimerBLAST will show a different melting temperature. PrimerBLAST uses the same algorithm as primer3, but has different default settings (which you can change, for example the Tm calculation method), but since it has a limited options I don't use it as a primary primer design tool, unless I want intron spanning real-time primers, which are difficult to design in primer3.
Can you post a pair of your primers from Primer3Plus you're not sure about (and the amplicon sequence), I can check if they're specific on BLAST with my very humble somewhat experienced opinion.
Dear Trof,
thank you very much for replying my posts.
I urgently had to order my primers, so I ordered the ones designed by PrimerBLAST and analyzed with NetPrimer.
This is one of the pairs with the amplicon (ST6Gal 1, Mus musculus):
gcc gtcgtgtctt ctgcaggatc tctgaagaac tcccagctgg
gtcgagagat tgataatcat gatgcggtcc tgaggtttaa tggggcacct acagacaact
''ga'' in bold indicates site of spliced intron sequence.
Sequence (5`→3`) Strand on template Length Start Stop Tm GC%
Forward: GCCGTCGTGTCTTCTGCAGGAT (+) 22 1338 1359 59.2 59.1
Reverse: TGGAAGTTGTCTGTAGGTGCCCC (-) 23 1444 1422 58.2 56.5
Product length: 107 bp
Total intron size: 6663
F primer: no hairpins, one dimer (ΔG = -11.24 kcal/mol), rating 79
R primer: no hairpins, no dimers, rating 100
FR primers: two cross dimers (ΔG = -6.09 kcal/mol, ΔG = -5.12 kcal/mol).
When I run this pair via Primer3Plus I get a message about too high Tm and high end self complementarity.
When I get them I can share my experience if you are interested.
Because of my work it is necessary to design primers spanning introns. You write it is difficult to choose such primers via Primer3Plus. I thing I found how to do this - it takes 30 min.
First,I go to the genomic sequence and write on paper the start and the end of mRNA, and the sites of splicing. I usually write on paper 3-4 codons before and after the splicing.
Then I paste the RefSeq mRNA in MS Word document, and use the function Find to mark the Start and the End codon, and also the sites of splicing.
Finally, I use small fragments comprising a site of spanned intron to choose primers in Primer3Plus. I always check the selected primers on the mRNA in Word to verify if they really span an intron.
Maybe this approach is stupid, I don`t know - everything what I know in the field of PCR is from this site and from the manuals.
I checked your primers in primer3, there seem to be some issues with higher values of ANY and 3', because as 3' is higher than maximum ANY seem to be within limits, but the message says otherwise. I encountered similar things with primer3 with primers designed elsewhere, I'm not sure if it's a bug or it has some other values that are not displayed.
Anyway, PrimerBLAST says it's fine, I would only be a bit worried by the high Tm of the primers, but that's something that depends on salts in reaction and things like that. So it's probably fine, all the tools just try to guess what primers are suitable, you will always see how they perform in real reaction.
As for the second part, I meant something else when I said it's difficult to design intron-spanning primers in primer3Plus. It's not difficult to design primers over single selected intron, but it's not possible to design primers over any intron and choose the best. If you have 8 exons, you can decide to design primer over intron 4, but you would need to design primers also over introns 1,2,3,5,6,7 separately and then select the best. You can say PrimerBLAST to just design over any intron and then choose the best.
Anyway if you are decided to design primer over one specified intron, writing sequence to a paper and finding it in Word is not exactly convinient. You can find your gene in Ensembl, which offers viewing in the Exon mode and there you simply copy one exon, put a < bracket before the last nucleotide, select next exon and put > bracket after the first nucleotide. Now you have the exon-exon junction marked as target and primers will be designed to span this two nucleotides. But as I said, you can only select one exon-exon junction at a time, but sometimes you just want intron-spanning primers and you don't care where they are located on the gene.
Only one thing is better to take care of, since coding sequences usually match between Ensembl and NCBI database, there can be differences in the UTR. In case of your gene there are two Ensembl transcripts that translates to same protein, but differ in the UTRs, and none of them matches the NCBI reference mRNA UTR, probably there are more variants transcribed. So it's better to select introns within coding sequence to design over.
And generaly, using non electronic stuff when working with sequences is laborous and can create mistakes. Copy/paste is your friend, and don't be shy to use notepad, (or better Notepad2) if you only need to store few parts of a sequence or search (you can use uppercase and lowercase letters to mark the beginning or end of some subsequence), Word is mostly unnnecessary for simple text operations and if you copy from it, you also copy the formatting, which is not desired usually. If you do use it, make advantage of highlights, different colors of text to make some extensive marks in the sequence (and better choose monospace font like Courier New). And as I found out today, if you display FASTA version of the sequence in NCBI, there is an option "Find in this sequence" that allows you to search through lines.