Dear Frank

• For general protein detection in a semi Quantitative sense Western bots are OK
• For Accurate detemination of specific protein levels using antibodies I would opt for ELISA (versus a standard calibration curve made from BSA)
• RT PCR would not be a good idea as their is not always a direct relationship between RNA level detected by RT PCR and steady state protein level
• I agree with the last commentator that MALDI TOFF would be the most accurate method for determining absolute abundance but I realise this is very specialised and you would probably need to outsource to detected persons in a proteomics facility
• If the protein is a surface membrane protein then flow cytometry might be an option but over all I would say the cheapest, easiest and most accurate main stream method utilising antibodies would be photometric signal from ELISA versus a BSA standard curve

To supplement what Laurence excellently wrote about detection of specific protein I think it is worth mention total protein assays as Bradford, BCA or Lowry, and their improved modifications. BCA is among the best assays providing the very accurate estimation if correct standards are used. Also, for a pure protein preparations very good idea is to use spectrophotometric UV assay and here the best, in my opinion, is this method:

Article An absolute method for protein determination based on differ...

• i concur the above: total can be easily accurately and very cheaply measured by a plethora of colorimetric assays like Bradford or Lowry by photometry in plate readers or by fluorometry
• How to quantify a single protein species you need be able to isolate. that protein with an antibody by immuno precipitation or better still ELISA where you can interpolate to a BSA protein standard curve