Hi Chen, you could try affinity chromatography. Isolate the small molecule, hook it up to a resin and see if the your target binds to the immobilized protein.

Thanks for your kind advice. I may first go with cellular thermal shift (since it's a label-free method), then if there is a positive signal, i'll try to immunoprecipitate the proteins, elute them from beads for further validation or attach a biotin to the molecule as you adviced. I wonder if there is a general method that could trace the molecule behaviors in cells, applying fluorescence or lumi for colocalization with the target.

Hi Chen, personally I would not go with thermal shifts because you will never prove that the shift is just caused by the ligand and not by exothermic or endothermic components in you ligand buffer.

I would say - go straight for IP. Nobody can argue with gels showing co-immunoprecipitation. (PS, I did not recommend biotin). 

This experiment should be very simple. Attach your ligand to AFFi-10 or CNBr Sepharose and do pull-down experiments.

Once you have convinced yourself and your critics that these proteins interact, then you can go crazy with biotin labels etc. 

Thanks so much for your kind suggestion.I quit the thermal shift.

However, I don't quite understand. Do i need to add primary amines group to my molecule, or how colud they hook up to the resin (since the manual suggest Affi-Gel or CNBr could react readily with primary amines)? 

Then if my molecule does has primary amines, what if it plays the key role in interacting with the protein?