Hi Inigio,

Can you describe what you are trying to achieve with the linker exactly? There are standard linkers that are used but may not be appropriate in your case. For example:

http://parts.igem.org/Protein_domains/Linker

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3726540/

http://www.sciencedirect.com/science/article/pii/S100200710900210X

The assembly question is secondary to your design question. Once you know what you want to build you can figure out how to build it. Gibson assembly is a very useful method along with many other similar methods that rely on homologous recombination. These methods suffer when the DNA you are assembling has repeats or very high or low GC-content. If your proteins and/or linkers have these features you may want to use a different method to assemble such as restriction enzymes.

I hope that helps!

Paul

I am assuming that a linker is needed when doing fusion proteins, is that preconception wrong in the first place? 

further details about the constructs and purposes:

ProteinX is a DNA binding domain to redirect the protein1 and protein2 respectively in each case to the DNA where they are supposed to develop their catalytic activity (gene1) and recruit other factors (gene2).

Interesting idea. It sounds like a linker in your case would be useful since you need the domains to be independent and flexible, although maybe this isn't the case. Unfortunately I have done only a few of these experiments using standard fusion linkers already published. I would suggest trying a few different standard linkers in parallel and seeing if any of them give the results you would like. A combinatorial assembly scheme can be done with both Gibson and Golden Gate assembly strategies. Programs like Teselagen (https://www.teselagen.com/) can help you design these assemblies and flag problems (such as repeats) that might be present in your designs.