Using BLAST to check primers - (Aug/11/2009 )
Hello all, I searched and couldn't find a topic on this (even though I'm sure it's there somewhere) but I have a question regarding BLAST and checking primers. So I am cloning a specific gene out of some gDNA from M. Tuberculosis and have designed some primers to do this. I was told I had to BLAST my primers against the genome to see if my primers were specific to the region I want to amplify. Well, I didn't do that at first, and sure enough my PCR turned out like garbage. I've tried figuring it out myself but have had no luck, so if anyone can help tell me how to use BLAST to check my primers, I would be more than grateful. I've attempted to use the primer design tool on BLAST but haven't had much luck with that either, plus, the primers I'm designed have to have bases complementary to the overhangs in my eK/LIC vector I'm placing the DNA in. Thanks!
Just load your primers and select appropriate stain in the following link.
http://insilico.ehu.es/PCR/index.php?mo=Mycobacterium
Hope this helps....
Thanks, that helped a bit, although I have finally figured out how to use BLAST properly. The primers I used got TONS of hits... whoops. I am confused on the "score" number though and if it is more important or just as important as the identity number. For example, a new primer I designed got a 90% identity and a score of like 74, however, there were other sequences which had 100% identity but only a score of 28. Is there a rule of thumb of what or what not to accept? Thanks again.
You're missing a piece of information -- how many bases comprised the 70% ID, and over how many bases did it match at 100%?
Also, are you using Primer BLAST?
HomeBrew on Aug 11 2009, 05:32 PM said:
Also, are you using Primer BLAST?
I've tried the primer blast but the problem is my primers have to have a certain overhang because of the vector I'm using. I have to have primers at each end of the gene for this. When I use primer blast they give me primers at various locations throughout the gene. I've never done primer design before and was kind of thrown into this on my own and am trying to learn as much as I can without having to ask for serious help from the postdocs around the other labs in the building. I'm getting different results now then what I was getting earlier... this is frustrating. So what I'm doing is going to nucleotide blast --> Putting my primer sequence in the "Enter Query Sequence" box and then putting in my organism in the "Choose Search Set" area. I then BLAST and get my "hits." I just got over 4000 hits for my primer, my target gene was first with a score of 42.1 (about 21/21 matches) and 100% identity, all the other matches were mostly 100% identity's but had scores of about 20 (matches of 13/13 or less). Less of my primer sequence matched them. Am I going about this the right way? It seems like I'm making this harder than it really is haha. Thanks for everything so far.
First off, it's good that your best match is what you expect, and that the match is 100% over the full length of the primer. I understand that you need to add restriction sites to the 5' end of the primers, but they won't participate in the amplification initially, so you should remove them from the sequence you're BLASTing as they'll just confuse the results. Use just the bases that match your target sequence. Secondly, even if you have a "good" match somewhere else in your genome, if it's off at the 3' end of the primer, it won't matter, because the primer will fail to amplify at this target location.
This second point is why I don't bother to check my primers by BLAST (BTW, I've designed and used thousands of primers). The value of BLASTing your primers is diminished by the fact that even if BLASTing shows a good match -- say 20 of 21 bases -- if the 3' base is wrong, the primer will fail. So there is a disconnect between a degree of identity match and what will really work in PCR, which really reduces the usefulness of performing this step.
The proof is in the pudding, as they say -- the way I find out if my primers are any good is to use them.
You said your PCR "turned out like garbage". No amount of BLASTing is going to fix that, so let's back up a bit. How did you design your primers? What were your PCR conditions? How much template did you use? Define "turned out like garbage".
HomeBrew on Aug 12 2009, 04:17 AM said:
This second point is why I don't bother to check my primers by BLAST (BTW, I've designed and used thousands of primers). The value of BLASTing your primers is diminished by the fact that even if BLASTing shows a good match -- say 20 of 21 bases -- if the 3' base is wrong, the primer will fail. So there is a disconnect between a degree of identity match and what will really work in PCR, which really reduces the usefulness of performing this step.
The proof is in the pudding, as they say -- the way I find out if my primers are any good is to use them.
You said your PCR "turned out like garbage". No amount of BLASTing is going to fix that, so let's back up a bit. How did you design your primers? What were your PCR conditions? How much template did you use? Define "turned out like garbage".
First off, thank you very much for the reply!
When I say my PCR turned out like garbage, basically I was amplifying the wrong region, I was getting a strong band around the 650 bases region whereas my desired product should be ~2.5 kb. The vector I'm using is a pET 46 eK/LIC vector, there are no restrcition enzymes involved in this vector, it is an open vector which contains overhangs in which my primers must have complementary base pairs too. The first 12 or so bases of both the forward and reverse primers are already "pre-chosen" if you will. So my forward primer sequence would have to start GACGACGACAAGATG..... and then into my gene. The reverse primer has to end with ...GGCCCGAAGAGGAG with having the base pairs of the gene at the beginning of that sequence. Initially I just took both of those sequences and went about 10 more base pairs into the gene and tried those primers, those failed as I previously described. I'm cloning out of the gDNA of the H37Rv strand of M. Tuberculosis, which is probably making this more difficult than if I were cloning out of a plasmid or something. If this turns out to be a problem with the vector I'm using, I may have to use another vector. My PCR conditions were as follows:
1) initial denat. for 10 mins.
2) denat. for 10 s
3) annealing: 15 s at 62, 65, and 68 degrees (3 different rxns)
4) extention at 72 for 1 min. 30 s
5) repeat 2-4 30x
6) final extension for 10 mins at 72
7) hold at 4
I used ~100 ng of template DNA and also 5% DMSO
Thanks again!
GACGACGACAAGATG looks horrible and unnecessary to me. 15 non-specific GC rich nucleotides is going to cause problems in a 25mer. you might want to design a 40mer by adding another 15 nt's to the 3' end of the primers...they should anneal more efficiently and the non-specific 5' end shouldn't anneal to your target in the first few rounds of your PCR.
what PCR mix are you using? plain old taq or a proofreading taq? a proofreader is ideal...something like Pfu turbo.
10 min. initial denature seems ok...might be a little long...i use 3-5 min.
try 30 sec. 94C denature.
try 20-30 sec. annealing but start at the high temp and proceed to the lower temp..touchdown pcr. starting lower will contribute to non-specific amplification. 68C is high and at this temp your taq is working just as well as at 72C.
depending on your taq...amplification time should generally be 1 min/kb. so you should be trying a 2 min. 30 sec amplification.
amount of starting material will dictate the number of cycles...35 seems a good start..but without a proofreader, extending the cycles may cause over-amplification of your product and increased error incorporation at 2.5 kb.
you can hold at 12-16C...it's better for your machine, especially if you're holding overnight.
the best PCR cloning vector i have ever used is called pJET (fermentas). the ligation is blunt but all your colonies have insert. your fragment when ligated into the vector disrupts a gene that when expressed in bacteria kills the cells. re-ligated vector prevents the cells from growing. the only chance for a false positive might come from cloning unpurified PCR product.
eldon on Aug 12 2009, 07:01 AM said:
what PCR mix are you using? plain old taq or a proofreading taq? a proofreader is ideal...something like Pfu turbo.
10 min. initial denature seems ok...might be a little long...i use 3-5 min.
try 30 sec. 94C denature.
try 20-30 sec. annealing but start at the high temp and proceed to the lower temp..touchdown pcr. starting lower will contribute to non-specific amplification. 68C is high and at this temp your taq is working just as well as at 72C.
depending on your taq...amplification time should generally be 1 min/kb. so you should be trying a 2 min. 30 sec amplification.
amount of starting material will dictate the number of cycles...35 seems a good start..but without a proofreader, extending the cycles may cause over-amplification of your product and increased error incorporation at 2.5 kb.
you can hold at 12-16C...it's better for your machine, especially if you're holding overnight.
the best PCR cloning vector i have ever used is called pJET (fermentas). the ligation is blunt but all your colonies have insert. your fragment when ligated into the vector disrupts a gene that when expressed in bacteria kills the cells. re-ligated vector prevents the cells from growing. the only chance for a false positive might come from cloning unpurified PCR product.
The GACGACGACAAG sequence is to code for an enterokinase site to where I can cleave off the N-terminal his-tag after protein purification. I've tried adding quite a few more nucleotides to the ends of the primers, but I start approaching a Tm over 80 before too long (at about 32 b.p.), I guess b/c it's so GC rich. I am using the Phusion High Fidelity PCR Mater-Mix kit by Finnzymes for my rxns. and am basically following their protocol for the PCR conditions, I'm not exactly which kind of Taq polymerase it uses. I usually take my rxns. out within 30 mintues after they're done, but if I ever leave overnight I will definitely keep the higher temperature in mind.
I have come up with this for the primers:
Forward primer: ACGACGACAAGATGACAGACACGACGTTGC
Reverse primer: GTCTGCCCGTTAATTGGCCCGAAGAGGAG
The bolded parts are the parts I must have for the vector overhangs, the non-bolded parts are from the gene itself.
Both have a Tm of ~79 C, but I'm still not so sure they will work too great for my strain. Maybe I'll just bite the bullet and order and try them.
Thanks!
NYYFan on Aug 12 2009, 10:12 AM said:
This is likely your problem -- you in essence designed a 10-bp primer, and I'm not suprised that the PCR looked horrible. What you need to do is design your primers first (I usually use 24 bp, but anything from 20 bp - 30 bp usually works well), and then add the 5' sequences you need on to them before ordering them.
Look at it this way. I don't use the vector system you describe, but I almost always add 10 bases to the 5' end of my primers (a six-bp restriction site and 4 irrelevant 5' bases to allow the enzyme to cut efficiently). These 10 bases are added *after* I've selected my 24-bp primers on the basis of Tm, secondary structure considerations, uniqueness of sequence, position on template, etc. -- right before I order them, I tack the ten 5' bases on.
How are you selecting your primers? If you're eyeballing them, I suggest strongly that you instead use a primer design program (I use Primer3). Design your primers without regard for the 5' sequences you need. When you have a pair that are a good match as to Tm etc., and prime where you need them to, and look good in BLAST (if you're going to do that), *then* add your 5' bases required by your vector and order them.
I ran the primer sequences you provided above without the 5' extensions. Here's what I got:
Primer pair 1
Sequence (5'->3') Length Tm GC%
Forward primer ATGACAGACACGACGTTGC 19 52.89 52.63%
Reverse primer GTCTGCCCGTTAATT 15 41.25 46.67%
Products on target templates
--------------------------------------------------------------------------------
>NC_000962.2 Mycobacterium tuberculosis H37Rv, complete genome
product length = 5631
Features associated with this product:
PPE family protein
PPE family protein
Forward primer 1 ATGACAGACACGACGTTGC 19
Template 2163152 ...C.G.G.C......... 2163170
Reverse primer 1 GTCTGCCCGTTAATT 15
Template 2168782 CA.AC.........A 2168768
product length = 2034
Features associated with this product:
long-chain-fatty-acid--
acyl-CoA dehydrogenase
Forward primer 1 ATGACAGACACGACGTTGC 19
Template 1509566 TG.G.G...G........A 1509584
Forward primer 1 ATGACAGACACGACGTTGC 19
Template 1511599 CCCGTT.........C... 1511581
product length = 5291
Features associated with this product:
hypothetical protein
isopentenyl-diphosphate delta-isomerase
Reverse primer 1 GTCTGCCCGTTAATT 15
Template 1976270 CCA.T.......... 1976256
Reverse primer 1 GTCTGCCCGTTAATT 15
Template 1970980 ACG..G......... 1970994
So, the Tm's of your forward and reverse primers are 52.89 and 41.25, respectively -- too low. Also, your primer sequences don't match anything on the Mycobacterium tuberculosis H37Rv genome completely. This is also a problem...