Two ideas come to mind. First if your molecule has an affinity tag, load a column with your protein of interest. Then take fractionated clarified samples from cell lysis and run them over the column. Wash off non-specific binders *carefully* cleave the tag (while in the column) and recover your molecule (hopefully) bound to it's partner. The second idea is a molecular biology techinque called the yeast two-hybrid system. I am pretty sure there are kits and protocols available to run this system. Good Luck!

Two ideas come to mind. First if your molecule has an affinity tag, load a column with your protein of interest. Then take fractionated clarified samples from cell lysis and run them over the column. Wash off non-specific binders *carefully* cleave the tag (while in the column) and recover your molecule (hopefully) bound to it's partner. The second idea is a molecular biology techinque called the yeast two-hybrid system. I am pretty sure there are kits and protocols available to run this system. Good Luck!

With the affinity purification (pull-down) method Bryan described, a useful control is to use the same resin without the known protein bound. This helps identify false-positives.

A typical follow-up to the pull-down is to run the captured proteins on SDS-PAGE, then identify them by mass spectrometry. This will, hopefully, yield a small number of candidate partners. Further work will be needed to demonstrate that one or more of them is a true partner.

I agree that the Y2H is a good start. Since you need a lot of tools you may consider to use a commercial service (e.g. http://www.hybrigenics-services.com/). I guess that you have already screened the databases:

http://string-db.org

http://thebiogrid.org/index.php

http://severus.dbmi.pitt.edu/wiki-pi/index.php/

pull down then mass, if the partner is in the library, you will know, normally you will get several.if the partner is new protein never been published, this will be a problem.

nice suggestion...