PCR Accuracy?

[BKallen]

If you are doing PCR with the intention of sequencing the product and checking a short region of the PCR product for a mutation, is there a region where taq polymerase is more or less likely to make a mistake? For example would it be more likely taq makes a mistake near the primers or in the middle of the PCR product? Should I be using a high fidelity enzyme even if it is only a short (20-25 bp) piece that I am concerned with?

[JackBean]

the mutation probability should be the same for each nucleotide. The polymerase doesn't know, whether is the preceding chain long 20 or 2000 nucleotides

[Papaver]

The total error rate of Taq polymerase has been variously reported between 1 x 10-4 to 2 x 10-5 errors per base pair. The smaller your product is the better. If you really want to be sure that your sequence is correct then take a polymerase with proofreading activity.

[BKallen]

So would it be more likely that a mutation occurs in the middle of the PCR product, near the primers or does it matter?

[JackBean]

as I said, the probability should be same.

[cw11]

I can think of one case where errors might be more likely to occur - when you have a large number of tandem repeats and strand slippage occurs. But I doubt this is going to be a problem for you.

In terms of PCR, if you are sequencing afterwards (as you say you are), the accuracy of most sequencing reactions tends to drop off sharply after about 600 nucleotides (but don't quote me on that, it's an off-the-top-of-my-head number) so you might see differences in error rate based on location because of the sequencing. But in terms of the polymerase itself, I can't think of a reason why the taq would me more likely to make an error in the beginning or the end of a transcript.

[JackBean]

you're right with the 600-700 bases in sequenation reaction, but that's IMHO relevant to separation than to amplification (the error rate is high also in the beginning)