There is little appreciation of the degree of validation performed by commercial antibody suppliers (especially large ones with catalogs of tens of thousands of antibodies). New antibodies developed by a PI in-house are usually greeted by an appropriate degree of skepticism by peer reviewers when they are first published. However, I believe the same is not true for an antibody from a commercial supplier, even if validation data has never been published. When such a company says their antibody is specific for a given protein, people generally take their word for it.
On at least a few occasions, I’ve ordered previously-unpublished commercial antibodies against a particular protein that I’m studying. At this point, I usually know some basic details about the protein I’m investigating, including which cell lines express it and its molecular weight. I also normally have expression vectors encoding the protein. Most of the time, even with antibodies advertised for western blotting, they simply detect nothing, even in cell lines which I know express the target. More disturbingly, in several cases, the supplied antibodies bind strongly to something that’s approximately the correct molecular weight of the protein of interest, but upon further investigation (by transfecting an expression construct or comparing expression in different cell lines), it is not binding to the right protein.
It is clear to me that most of the hundreds of antibodies these suppliers add to their catalogs each month receive only a cursory validation. I guess when a PI purchases such antibodies, they are essentially paying the company for the honor of validating their products!