Aggregation Index Experience? - (Dec/12/2013 )

Does anyone have experience with using an Aggregation Index (AI = 100 x (A340/(A280-A340)) to monitor protein aggregation?  I've pulled up some papers where it has been used but it seems like the A280 would need to be large (>5) and the A340 small (<0.2) to have the AI fall bellow 2 and be read as soluble and my A280's are rarely that large.  Was hoping someone had some practical experience.....

I simply measure A320 and A280 and calculate "% aggregation" as = (A320/A280)*100. Usually a % aggregation of less than 2% is great and anything above 5% needs attention.

Of course these cutoffs are arbitrary and depends on the protein and its final use/application.

Thank you for the reply.  Perhaps a silly follow up: where my A280's typically fall in the 1 - 4 range, to obtain a '% aggregation' lower than 5% the A340 would need to be about or less than 0.1.............isn't this uncomfortably close to the confidence/variability of the spec?  I'm worried my readings will be in the noise.

Yes that's true. I am comfortable making measurements between 0.05 and 1.2 on a UV spec with 1cm path length. I get the A320 background of my blank buffer separately ( don't blank the spec).