Need help with 7-way ligation. - Help make 7-way reliable cloning strategy! (Jul/30/2010 )

Desperate! Have been doing this 7-way ligation all week and it has not been working. Have screened 288 colonies but no complete plasmid found. 7-way has worked twice before. What am I doing wrong now?. All ends are incompatible.

Okay, enough joking around.

I have been attempting to see what is limit for multiway ligation aided with a 12-multichannel pipette. I have found that 6 way ligation is possible and thus far has been reliable. However I have noticed a sharp drop in efficiency at 7 ways. 8-way has never worked. It is probably too much to get an answer of why that is so. Thus i will move to a marginally easier question, "Are there any ways to improve ligation efficiency?"

It is certainly possible to screen more colonies. Depending on colony pooling format, I can have a maximum screen of either 768 colonies or 1152 colonies per 96 well PCR plate. Colony PCR on pools is a possible brute force approach that has reduced use of reagents. But the time to pick 768 colonies or more would be substantial. Thus i am looking for alternatives.

Are there any little things you do to improve the efficiency of accurate ligation?

Or is this the limit of staggered end ligation? There is homologous recombination based plasmid construction method (similar to SLICE) that is sold by Clontech.

EDIT: Hmm here is a thought, replace T4 DNA ligase with NEB E.coli DNA ligase. NEB says E.coli DNA ligase is more specific but caution against using it for cloning. I wonder why?

EDIT2: Well I guess the next obvious thing would be to dephos every alternate DNA fragment. This would reduce self ligation and further reduced the number of possible ligation combination (under conditions of reduced fidelity) that would result in a circular plasmid.

For ligating dsDNA fragments with cohesive ends, either E. coli DNA ligase or T4 DNA ligase can be used. T4 DNA ligase can join blunt-ended DNA fragments. E. coli ligase is unable to join such fragments unless PEG is added.
E. coli DNA ligase may be used to ligate the sticky ends of two DNA fragments together, leaving the blunt ends of the fragments available for a subsequent insertion into a vector.