Not working in some samples? If they work with some samples it is not a problem of the primer pair I guess. Have a look at the samples you have problems with. Are they out of the same RNA-isolation you did? Do they have the same condition? There might be something, which inhibtis PCR. Well, I guess you have checked for a housekeeping gene. You might want to get a positive control (some cDNA from cells or organs you definitely can be sure of that they should contain the product). 

Otherwise, if you decide to order new primers, be careful that you keep GC around 50% (well, not more than 60%) and 3'/5' selfalignment as low as possible. I dont know how you look for your primers, but a good website is pubmeds' primer-blast:

https://www.ncbi.nlm.nih.gov/tools/primer-blast/index.cgi?LINK_LOC=BlastHome

Good luck - and PCR

Sure, you try different primers. Generally, shorter primers increase binding efficiency. You can also use a pair of primes to amplify the fragment containing your target, and use the resulted product as a template to amplify the target use your previous primers.

I agree with Wenguo Cai, when a primer is short the possibility to find the complementary sequence is higher than long primer.

Try to see the principle of Nested PCR. That may help you.

All the best!

I too agree with Wenguo Cai Aa shorter primers increase binding efficiency . you can check your primer may be its length is inhibiting binding it also depend on whether your protecting DNA from evaporation in PCR  and PCR . if DNA is Damaged Amplification is not possible . repeat the steps carefully

Not working in some samples? If they work with some samples it is not a problem of the primer pair I guess. Have a look at the samples you have problems with. Are they out of the same RNA-isolation you did? Do they have the same condition? There might be something, which inhibtis PCR. Well, I guess you have checked for a housekeeping gene. You might want to get a positive control (some cDNA from cells or organs you definitely can be sure of that they should contain the product). 

Otherwise, if you decide to order new primers, be careful that you keep GC around 50% (well, not more than 60%) and 3'/5' selfalignment as low as possible. I dont know how you look for your primers, but a good website is pubmeds' primer-blast:

https://www.ncbi.nlm.nih.gov/tools/primer-blast/index.cgi?LINK_LOC=BlastHome

Good luck - and PCR

Is there any known difference between sanples for which the PCR is working or not? This is the first clue to the reason of the problem, because primers are the same for all the samples.

Anyway, if you are going to redesign the primers, take into account not only the lenght but also whether they can form hairpin loops, bulge loops, dimers and other undesired structures at temperatures close to the one you use for PCR annealing. Many softwares give you that information.

Dear Somin

• Primers if specific should always find your sequence !!
• Generally primers are between 18bp and 25bp conducive to a Tm of 55C to 65C
• DO NOT go below 18bp (what do you mean by 'short') as they will then start to bind to secondary non specific sites
• I have attached a document for good primer design
• Generally, if your primers are between 18and 25bp and they do not amplify then either priming efficiency is poor owing to poor priming design; your primer does not match your target sequence, particularly in the last 3' 5bp or your annealing temp in your PCR is too high (should be about Tm-2C)
• Have you blasted your primers to make sure they perfectly match your target and only your target ?
• Do your primers end in either a GC; GG; or CC residue to ensure efficient priming and strong PCR ?: At the very least do they end in either C or G or CA; GA; CT or GT ? The former double purine endings are preferable but the latter combinations are also compatible with efficient priming
• Do not end your primers in double pyrimidine residues; that is TT AT AA or TA. This will increase your chances of poor priming efficiency
• Thus, if your primer is longer than 25bp it can be reduced in size down to about 18 bp but as mentioned by Teresafor this size range you require 40-60% GC content in your primer; also for efficient priming but also to make sure that your TM or primer melting temp is > 55C and < 65C such that your annealing temp in PCR - usually TM-2C is 55C to 65C
• Generally,  you can either situate a GC rich region or GC clamp in the middle of a primer or if not terminate your primer with a GC; GG; CC; or CG residue
• Avoid making your 5' region GC rich as it is the 3' half of a primer and in particular the last 5bp that confer strong specific priming
• See attached

if some samples work and others do not reproducibly then my first 2 guesses are that the dna quality is poor in the ones that fail or that the quality is good but the 3' end of one primer sits over a polymorphism so there will be no amplification. In this case move both primers out by 10 bases or more. You could move them inwards but then you will not know if the polymorphism exists when you sequence the product

If some samples are working means, It is not the problem of your primers. It is the problem of your other samples DNA quality.. Check the quality of yur DNA samples

It could be a polymorphism at the place of the primer, or, depending on which region you are sequencing and the particular group of fungi, you could have an intron that is inserted close to the primer site in some samples that is disrupting amplification (not usual in ITS but common in some LSU and SSU). Choosing a primer site that is away from (upstream of) the one causing problems often works in either case - even if there is an intron, if it is far enough away from the primer site, amplification will it will often 'read through' it. If the DNA has loops in it, that will also disrupt amplification, in which case, you could try using DMSO before extraction to 'relax' the strands, but in some families (e.g., Entolomataceae), it causes the strands to fall apart into short segments.

I think there are a lot of good answers here. Sometimes primers don't work well at the temperatures of the calculated TMs due to secondary structures, heterodimers, etc. One thing I always do before ordering new primers is a gradient PCR. This almost always works at finding the right temperature for me. If your thermocycler can do a gradient I would start there. 

Gradient can also be done in a real time cycler, is may sound like overkill, but works well.

I recommend you would try to run using touch down!

if not working in some samples than primers seems ok. your samples might have some problem. you can also check expected band size and specificity of primers using primer blast option in NCBI. that will give you an idea is there any other binding site for the primers or not. but it seems from your question that primers are perfect but samples having problem.

Do you think your primers are "not working" due to genetic variability in your samples so that your primers are no longer annealing to the template?  If so, you may need to do an alignment of all of your samples in your region of interest and design primers that anneal to the most conserved region.  Or you could design degenerate primers for that region.

Do you have a good positive control PCR reaction that works for ALL of your DNA samples?  Something like an actin gene that is highly conserved is a great control.  If you are confident that you have quality DNA from all of your samples, that helps narrow down the trouble-shooting.

Also, as a rule of thumb, longer primers are more specific, shorter primers are less specific.

Yes, you can definitely go for longer primers but in my, opinion you should first cheek your template DNA, dNTPs, and enzyme that whether they are working or not and you can be optimized your PCR program. 

I think your primers are ok because some samples are being amplified. May be their is Problem in the other sample... Load 2 UL of each of those sample in agarose gel and notice are they coming out from the lane..if so their must be ethanol/phenol contamination. If it ok then check the DNA concentration because very high or low concentrations positively hinder the amplification process. If it is also ok then once resuspend the purified DNA in nuclease free water rather than TE. Because sometimes ionic strength of the buffer could hinder the amplification.. adjust the primer concentration in respective to DNA..because at low DNA concentration the high primer concentration may produce primer dimer and hinder PCR process.

Just keep all the necessary condition (GC content , tm etc) very close to your heart and consider aligning the gene sequence prior to designing your primer and make use of only their conserved region.  Again, the fact that some of the old primers amplified some sequence is only pointing to the genetic variability of your organisms. Hence the reason for non amplification with others. Best wishes 

If your PCR is not working 'in some samples' then the problem is likely to be the samples, not the primers. 

In order to check this, you can do a spiking experiment. Repeat the PCR with the negative sample and a small amount of added positive control (about 1000 copies) in the same tube. If the result is now positive, your sample probably doesn't contain any DNA; if it is negative, but the positive control w/o test sample is positive, then the sample is inhibitory.

Note that this only detects differences in sample quality. The problem could also be due to sequence variation near the 3' end of the primer.

As many of the other respondents noted, using a primer-design program makes sense. Primer3 works well and it's free. I use PrimerExpress from Applied Biosystems, but unless you get it free with an instrument, it's very expensive; but it does make good primers.  

You have to change the primer design if it's not working for all the samples. However, if the PCR does work for some samples, this means there might be problems with your samples or the PCR. Therefore, Check the samples that didn't work. If the samples are fine, Optimize your PCR program as it might have something which impedes the PCR work.

Dear Somin. I'm not sure why I didn't pay more attention to the fact that some times your samples work and others not. What I have said above is good advice and well designed primers will give you more consistent data. Nevertheless

I can provide more technical details if you require 

First of all u need to check the way u r designing primers is right.

Secondarily these term "long" and "short" are somewhat arbitrary.

If u r sure of primer desigening then u must keep the length of your primer somewhere 20-28 bp with a Tm ranging in 55-62 C. GC content is another aspect, u need to pay some heed on and it should be no more than 55%.

Hope success will strike you!

Before you design new primers, make sure that you have good quality DNA from all of your samples. Try amplifying a gene with a pair of primers that you KNOW always works (e.g. actin, CO1, etc.). That way you can see if the issue is the primers, the DNA, or some combination.