PCR cloning - how many clones must one sequence?

Wolfgang Schechinger hubahopp at gmx.de
Thu Mar 25 07:13:18 EST 2004

Dear Bo,

Proofreading pol is a must, I'd say. Taq can make up to 1 error per
kilobase (depends on the number of amplifications of a gene and the
reaction conditions), so if you need accuracy, it might be worth
sacrificing some Krones more for Pfu, Pwo or smilar enzymes.

Normally, when I have sequenced one clone of a single isolate, and it
matches the prediction, I tend to believe it. However, I never trust just
the text file that comes from the sequencer, but always scrutinize the base
plot. Depending on the source material, if you have heterocygous genes, of
course, you might obeserve polymorphisms, though.
If you expect polymorphism, then it might make sense to sequence 5 or 7
clones (ask your favourite statistician for advice) or better use a
different method like real time PCR with appropriate probes for detecting
homo- oder heterocygosity.

Invitrogen (no affil, I just talked to them accidentally on a trade fair
yesterday) offers a polymerase formulation that includes single strand DNA
binding protein to improve selectivity in nested PCRs. 



At 11:18 25.03.2004 +0100, Bo Johansen wrote:
>We are cloning products from a nested PCR and single base variation 
>occur in many clones. We use degenerated primers and taq polymerase for 
>both PCR reaction. How many "alike" clones do we have to sequence in 
>order to be sure that we have the correct sequence?
>Will it help if we use a proofreading enzyme instead of taq?

Wolfgang Schechinger


More information about the Methods mailing list