Other than Taq? - for PCR after RT (Feb/15/2007 )
do you know which one is the best to do PCR from cDNA of 3Kb?
The Taq doesn't seem to be efficient enough in this case!
If 3kb is the total size of the cDNA, then any high fidelity polymerase would do.
KOD hifi, Vent etc...
You don't need the really long template polymerase like phusion etc.
High Fidelity polymerase is very expensive though compared to Taq.
Ah but it will save you from falling to your knee screaming "Nooo!" when you find that your sequencing results show that nearly every single isolate you've sent has a basepair mutation somewhere.
Taq has an error rate somewhere around (1 x 10-4) to (2 x 10-5) errors per base pair.
Assuming (2 x 10-5) as the error rate per base pair
running 25 cycles
The probability of an error in the fragment
= fragment lenght * error rate per basepair
= 3000 * (2 x 10-5)
Probability that fragment has no error
Probability that after 25 cycles fragment still has no error
Thus best case senario, only 1 in 5 fragment will be error free (21%). Worse case senario, the number is down to only 4%.
I would say it is definately better to spend the money buying a proper proof reading enzyme (avoid buying any "proof reading Taq"... the fidelity is better, but not by much.), it will save from trouble later.