I am troubleshooting a PCR reaction that gives me no product, and there are a lot of different approaches that I can find in books and just in google searching (things like running a gradient, increasing cycle number, etc.). But it would be really nice if somebody could list out in order the steps they take to troubleshoot a PCR reaction. Thanks.
I think 30 cycle PCR is maximum to see bands by electrophoresis in most of cases, because under the ideal efficiency, one copy could be detectable by 34 cycles of PCR. However, if your expected products are longer than 1 kbp, you might extend several cycles.
I usually add DMSO 5% in any PCR reaction. In my experience, DMSO would not interfere PCR reaction--you can amplify your fragments without DMSO; you could get PCR products with DMSO. Betaine sometimes work better than DMSO, so I am shifting my default reaction mix from 5%DMSO to 5%DMSO and 0.5M betaine.
For annealing temp, you can not tell which is better, lower or higher than the first trial which showed nothing. When increasing the annealing temp, primers may be able to access your target sequence more and you could get efficient amplification. So you might want to try both. If you have a PCR machine set various temperatures at once by gradient. You could include lower and higher temp together.
Alternatively, I sometimes use the touch-down method. You do not need to determine the best annealing temp in this case. If you just want amplified products, this could be a quick solution.
Additionally, hot start is also my default condition, because this could not ruin PCR reaction without hot start.
Basically, I change annealing temp. If I do not get good amplification and specificity, I change PCR enzymes. Changing PCR enzymes could mean changing buffer conditions because companies try to give unique features for their products.
Generally I start with a temperature gradient running in 2 deg steps above and below the calculated Tm (which I calculate by the simple 2AT4GC method). If that fails, I might try something similar along with PCR "enhancers" - usually some variation on DMSO or betaine. Depending on the problem (detection/cloning from a reverse transcription reaction? Sequencable fragment from genomic DNA?) I might try redesigning the primers, changing the type of PCR polymerase, or improving the template in some way.