I'm planning on creating a RADseq library, which naturally involves digesting genomic DNA following P1 adaptor ligation. I've read a few protocols that require any over-hangs resulting from restriction enzyme activity (e.g MspI or Nlaiii) need to be repaired (blunt-end repair) followed by adding a single A nucleotide at the end so as to make the DNA fragments "sticky". But is this really necessary? Wouldn't it be more efficient to custom make P1 adaptors so that they have a complimentary 3 or 4 bp overhang specific to the restriction enzyme used instead?
What I want to ask is what are the practical advantages of having a single nucleotide overhang vs. 3 or 4 nucleotides?
Thanks in advance.