Gene construction advice needed

"David Alexander dalex at
Mon Jul 10 11:47:45 EST 2000

Joseph C. Bagshaw

You're overlapping oligo strategy might work...

There are groups trying to solve math problems with PCR.
There was an article in Science a few years back:

Science 1994 Nov 11;266(5187):1021-4
Molecular computation of solutions to combinatorial problems. 
Adleman LM 

They were trying to solve the 'salespersons' problem.
i.e. finding the shortest distance between a set of destinations.
without visiting any destination twice.
(it's an easy problem if there are only 4 or 5 destinations,
 but with 10, 100 or 1000, it becomes very complicated
 especially  if there are limitations on any of the destinations
 i.e. you can go from A to B, but not from B to A)

To make a long story short, they created an oligo for
each possible trip - but designed them in such a way
that each destination had an unique  sequence.
They threw all of the possible oligos in together,
amplified them, separated them by size,
and isolated the product of the appropriate size 
to give all destinations, without visiting any twice.
(i.e. if you had 7 destinatinos, the fragment would be
 the length of 6 oligos - or maybe 7 if you had to
return to the first destination....)

In any case, combining oligos might work to build up your sequence,
the problem is producing overlapping ends that are exclusive
and yet, still give you the gene sequence you want.

Good luck,


dalex at


"Joseph C. Bagshaw" wrote: 

> I'm looking for advice on a gene construction project. I'm trying to
> assemble a 219 bp gene from oligonucleotides. I designed the oligos
> so that each overlaps the next by ten complementary bases. The plan
> then is to fill the gaps, then ligate.


More information about the Methods mailing list