Posted 23 January 2012 - 11:22 PM
Posted 24 January 2012 - 02:09 PM
Posted 25 January 2012 - 06:23 PM
Thank you very much, bob1.
Absolutely - check out things like yeast two-hybrid systems where you have a "bait" protein and a "prey" protein, I would estimate that over 50% of the interactions identified during these screens are meaningless or non-specific.
Indeed I got that interaction by yeast two hybrid and later confirmed the binding by IP. I didnt do IP in endogenous level, which looks like a big mistake now because those proteins might not be localized together. As what you have said, that interaction might be false-positive. And I read from Wiki and it is pretty clear that regarding results of Y2H, there is a lot to doubt about.
"For example, overexpression can result in non-specific interactions. Moreover, a mammalian protein is sometimes not correctly modified in yeast (e.g., missing phosphorylation), which can also lead to false results. Finally, some proteins might specifically interact when they are co-expressed in the yeast, although in reality they are never present in the same cell at the same time."
OK now I am clear about my proteins but I still want to know if at endogenous level two proteins bind or are associated, is there a chance this interaction is meaningless or by accident?