Absolute qunatification? relative quantification? compare different proteins to one another? big difference there.

Absolute quantification at this stage.

I don't think absolute quantification of the whole proteome is feasible (economically). you have to get isotopically labeled peptides for each and every protein you're interested , produce calibration curves to ID the linear range of detection for each one and then spike them into the original mixture and then run again. 

on the other hand, if you only want to identify how protein X changes under the different conditions, then you simply run the data through a quant software fitting for DIA, like DIA-Umpire

Cheers,

David.

Thanks David. I agree that absolute quantification of whole liver proteome is a herculean task requiring labelled peptides. How can we address the issue of experimental error while performing relative quantification? Do you think it's a good idea to add some exogenous proteins and look for their recovery? We want to make sure that any relative change which we see is due to the condition/treatment and not the experimental/operator error.

Hi,

you have to define what you are looking for to begin with (what is the exact question), and how you performed the experiment, for me to be able to advise you more specifically. is it a label-free quantitation experiment? were all samples prepared from the same type of tissue? same day? same reagents? same person? did you run all experiments with the same quantity of protein? did you set up the queue on the MS in a way proper for such experiment? how did you search the data? 

You may not want to spike a known protein if you fear that you're going to supress other proteins (although 50fmol of BSA should be too much), but at the very least, process a standard protein mixture as a control to make sure that your sample prep is good and that the instrument is calibrated well.

cheers,

David.