What's the typical route taken to obtain synthetic DNA/RNA samples in practice? Are they pretty much always custom ordered from a place that prints them using sold-phase synthesis, or can you cobble them together from random "common" samples?

Theoretically I could see an arbitrary 120bp oligonucleotide being made of 2-3 strands from random parts of across a genome or two, where you could separate out those parts with restriction enzymes and then anneal the snippets, but A) I don't know how you'd find those snippets reliably or whether my assumption that they'd exist in a random sequenced genome, or B) whether it's even efficient to try to digest, isolate, assemble and purify oligos that way when compared to buying a printed synthetic oligo

  • 2
    $\begingroup$ Depends on your application. For smaller to medium sized DNA oligos (<~120 bp) or medium sized dsDNA fragments (<3kb), I will always order them. What you're describing sounds a bit like Gibson assembly cloning (en.wikipedia.org/wiki/Gibson_assembly), and I would only go that route for fragments upwards of ~3kb, unless I needed to make several of them at once (also assuming I know the full sequence needed). Larger fragments can be purchased, but the price per bp can get high for very long fragments (still might be worth it for a one-off). $\endgroup$
    – MikeyC
    Oct 2, 2023 at 17:01
  • $\begingroup$ I see, is the same true for RNAs or would you just order a template DNA and run in vitro transcription? $\endgroup$
    – Gumpf
    Oct 2, 2023 at 17:22
  • $\begingroup$ @MikeyC i think that your comment is a different (and possibly more useful) answer from mine if you want to post it there $\endgroup$ Oct 2, 2023 at 17:44

1 Answer 1



For short defined sequences, IDT or a similar oligo synthesis provider is your best bet. It's very fast and straightforward, compared to trying to do something complicated with a genome. Recall that the number of possible e.g. 120mers is $4^{120}$, or $1.7668471*10^{72}$. That's an astronomically large number, and the probability of finding it in some random convenient genome is small.

For e.g. single well-studied genes, you might order a plasmid vector from AddGene or similar, which you can put into E. coli and grow up to make many copies.

If you are trying to isolate something a little less standard, you will usually clone it out of the genome though.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .