The secret, as in so much of science, is repetition. Never trust a single Western, but several showing the same thing might be construed as actual data -- particularly if conditions can be varied (e.g. treat cells with different drug that has the same target, probe with antibody from a different source) without breaking the result. :-) - Bill Hooker

In fact, I probably wouldn't trust most scientific experiments without x number of technical replicates (x depends on the type of experiment) - Deepak Singh

What do you think of attempts to 'standardize' Western blot to generate quantitative data? eg <http://www.ncbi.nlm.nih.gov/pubmed...> - Thomas Lemberger

What always makes me smile is people in cell biology saying how hight-throughput experiments are just noise when so many western blogs end up in the garbage because they don't look right :). At least the high throughput methods usually comes with a some sort of quality measure (accuracy, coverage, false positive, etc) - Pedro Beltrao

Thomas -- at a glance those look like good ideas; in practice few labs will adopt them because it will mean a lot more work. I've considered randomizing loading order on some of my blots, but I'm always so pressed for time and under such pressure to generate "figures for the paper" that I've never actually tried it. The best I currently do is not to over-sell blot results, and only use them as evidence for fairly robust effects. - Bill Hooker

Pedro - (I like the geeky lapsus "Western blogs" :-) I think you are right. On the other hand significant knowledge has been accumulated with good old Western blots. I guess as long as qualitative results are required it is fine. It is indeed different when data is meant to be quantitative, HTP or shared and integrated. Not sure how to deal with these classical mol biol techniques... We have to rely on trust... - Thomas Lemberger

ELISAs FTW, baby! - Mr. Gunn

Back in my wet lab days, the western was one of my favourites. Mine were always lovely. The secret is good antibody, cheap supermarket skimmed milk powder and an old-fashioned semi-dry blotter. - Neil Saunders

Thx guys for the knowledge sharing here. I learned a lot. - Attila Csordas

@Neil: if you have a good antibody you can dilute it in yak spit and your blots will come out great. The problem is, most antibodies are terrible. Bring on the aptamers already! BTW, anyone know anything about this: http://www.sysbio.org/datares... ? - Bill Hooker

Yeah, this reminds me why I quit the wet lab :) - Neil Saunders

A saying in our lab (passed down two generations of PIs) is "If you don't have a good antibody, you don't have a good project" ... needless to say, we end up expending a fair amount of effort raising antibodies (which don't always work). Sometimes a bleed will work on SDS-PAGE, but not on a native gel, which is a real PITA. At the pub last week I heard several a stories about completely useless commercial antibodies ... including the companies which have clearly fabricated their example blots. - Andrew Perry

Yep, bring on the aptamers !! While we are on it ... anyone know a decent commercial His6 antibody ? Most are rubbish ... which reminds me I've got to get around to trying that HisProbe (NiNTA-HRP) stuff from Pierce ... - Andrew Perry

Andrew, I've only had luck with His-NI affinity tags in a column format. - Mr. Gunn

Eva: we call 'em Santa Crap around here, but honestly I don't think they're any worse than average. Most companies sell Ab they know are weak at best. Something like BioRoot (now defunct) might have been able to provide a community recommendation system.... What I do is a google image search for "protein of interest" "Western blot" (or IHC, or whatever) to look for published uses of antibodies. - Bill Hooker