Basic Subcloning Question

George Rutherford gruther at
Tue Jun 1 13:30:51 EST 1999

In article <7j0td0$sfp$1 at>, chrisb at (Chris
Boyd) wrote:


> : I'd *strongly* recommend dephosphorylating your vector and redoing the
> : ligation. While it's true that you only need one good clone, finding that
> : clone would be the hard part if you forgot to dephosphorylate the vector.
> : Thermodynamically, reclosure of the vector is favored so much over
> : reaction with your insert that, unless you get incredibly lucky, you'll
> : have to screen hundreds (more likely many thousands) of colonies to find
> : one with your insert. Worse, you could screen every colony you have and
> : still end up with nothing but religated vector. That's a lot of work for a
> : low probability of salvaging the experiment.
> Where did you get that idea?  It's simply not true at all.  I have done
> scores of sticky-end ligations without phosphatasing the vector and get
> positive clones at a frequency of up to 70% on occasion, with the range
> usually being from 5-50%. Reclosure of vector can be minimized to < 50%
> if (a) fragment size is as small or smaller than the vector, (b) the
> vector:fragment ratio is optimal and (c) the concentration of ends is
> sufficiently high. The maths demonstrating this is complicated
> (although if you think about it you will see it must be true), but the
> empirical evidence is overwhelming: phosphatasing did not come into
> common practice until many years after cloning started but people still
> managed to come up with the goods.

Quite right. Years ago, when I was a tech, I worked with a postdoc who was
convinced (rightly or wrongly) that phosphatasing the vector was causing
problems with ligation, so we just screened by colony hyb. Worked every
time, and, frequently, more than 50% of the colonies were positive.


More information about the Methods mailing list