**3.2 Conventional methods**

There are a number of quantification methods for proteins either bioassays or instrumental methods. The most commonly used methodologies in bioassays are the Biuret (Savory et al., 1968), Bradford (Bradford, 1976), Lowry (Fryer et al., 1986), and bicinchoninic acid (BCA) (Smith et al., 1985) assays. These methods employ chemical reagents which specifically react with proteins to produce colored products which can be measured by UV spectrophotometer in a concentration-dependent manner. The absorbance of colored sample is compared to standard curves constructed with a known protein (frequently bovine serum albumin) in order to determine concentration in unknown samples. Biological methods are widely used owing to their simplicity and low cost, but are hampered by poor accuracy and reproducibility since their responses are either assay or calibrator-protein dependent, and also the results can vary with the residue composition of the target protein. Thus bioassays lead to relative differences in protein quantification and cannot give absolute values. On the other hand, instrumental analytical methods including chromatographic techniques (HPLC, GC), capillary electrophoresis (CE), and mass spectrometry (MS) are often favored because of higher precision. These methods may require time-consuming sample preparation, which leads to higher costs than biological methods. Nevertheless, both bioassay and instrumental analysis need to be calibrated with a highly reliable protein standard material to ensure accurate and comparable results. A higher order analytical method needs to be established for accurate determination of such protein standard materials.

The same is true for high throughput analysis formats such as 2D-PAGE gels (Smithies and Poulik, 1956) and enzyme-lingked immunosorbent assay (ELISA). These methods are particularly well fit the purpose of high throughput analysis for proteomics research and protein-chip analysis, respectively. Therefore, they have become powerful tools for screening effective protein markers. As such research progresses, it has become obvious that reliable quantification in an absolute manner is essentially required for comparability of data.
