If this is a general sequence there may be clones containing your sequence available. If it is for a specific individual it would be technically easier to amplify in smaller overlapping sequences. If the next stage is sequencing then NGS may be an appropriate option

Dear Leonid Ouchanov

As Paul also mentioned the best option is using overlap extension PCR, which first need to amplify short sequences of your whole fragment which these small sequences have an overlap regions with each other (overlapping at the end of one to start of the next) and in this way you could assemble small sequences together and make a big fragment. It is a bit tricky technique, specially when you want to assemble more than 2 sequences together. But another important point for amplyfing such a big fragments is that using a high fidelity polymerase such as "Q5 Neb" for all steps is suggested, using Tag polymerase most probably could cause some errors during amplifying such a long fragment.

Thanks! I do have primers to make shorter sequences, all I need is to choose polymerase. And yes, this is for NGS.