Too true. I can't tell you how many papers I've reviewed whose main conclusions relied on staining that was shitty. It's hard to just say "I don't believe you" in a review, so you have to come up with something nicer, such as, "The images seem overexposed, and the morphology of cell X seems peculiar. Perhaps some negative and positive controls would be instructive." Seems like biology 101, but the frequency with which it happens is astounding. And I assume that a lot of these papers get through, because the reviewers aren't experienced microscopists. Although I'm sure I miss a bunch of basic shit that other scientists would look at and say, "are you stupid?" If I only knew what those things were...Now I'm entering the realm of staining with antibodies and I'm learning just how few good antibodies there are compared to how many proteins I find interesting.
Yeah, considering how ubiquitous antibody-based conclusions are in biology, I'm surprised there's not a more thorough creation and validation process. I straight up do not trust polyclonal antibodies (it's not a defined reagent!) and vendors have almost zero standards for showing specificity. I'm looking at ~20 targets to validate, and even if I could find antibodies for them all (many do not have mouse-specific versions), the process of formally checking antibody would be close to half a year of work. If I have my way though, "Mass Spec" will be Aim 1 of my thesis proposal, "Staining and Validation" will be Aim 2, "Drugging" will be Aim 3, and "Cell Culture Follow-up" will maybe be Aim 4. I'm hoping the results are exciting enough that doing the first two and doing them well will be sufficient for my (hypothetical) committee.Seems like biology 101, but the frequency with which it happens is astounding
We lately had an interesting target for a mass-spec screen we wanted to follow and found out the THE experts for this protein sit right next door. We had a meeting, they told us everything they do, they have a knockout mouse (yes!) and a good antibody (double yes!) that we could use. We went back to the lab and tested the antibody on our tissue of interest, no signal... Lets test the antibody on kidney tissue, it has been published that it is expressed there. We see a signal! It is different than what was described, but it is a staining. Lets check in the knockout mouse, just to make sure. We see the same signal as in the wildtype. We test a few more tissues. Nope. Their antibody that they rely on for every staining they are doing is unspecific, sticks to everything that has a calcium-dependent lipid binding site and they never thought about making sure it is specific in their knockout animals that they had for yearssssss. WTF? We learned our lesson. Every antibody is tested and then confirmed with an in-situ to see colocalization before we trust it. thundara congrats on seeing the thing you wanted to see :)
Jesus. Well a least you had all of the resources on hand to invalidate the antibody ;-) Coincidentally, the department's new hire who is the door over from my adviser worked on <thing I'd been looking for>. I'm on the fence as to whether I want him or a more neuro-y person evaluating my work (or both???).
I think that having two different views on your matter of interest is very nice! They tend to give you very different but useful insight on how to tackle your problem. Had a similar thought about who to take for my thesis advisory committee. I wanted to add someone who is not in neuro, but my boss doesn't think it is useful. Maybe for next year :D