generating terminal truncations of a protein

[ChenHA]

Artem Evdokimov wrote:
> 
> There are
> also randomized PCR-based methods for generation of random truncations on
> both ends, these require the user to separate the resulting clones and
> analyze them by sequencing in order to establish their identity. 

If you have a reference for that it would be much appreciated.  I may
have to try that method some time later to find a soluble domain.