Does your sample consist of purified virus in a solution that contains no other proteins? If so, then you might be able to use one of the commercial protein assays kits, such as Bradford, BCA, or Lowry. If the sample is too dilute, you can concentrate it by ultrafiltration.

If the virus is in a solution that contains a lot of other proteins, such as medium with fetal bovine serum, then you will either have to use another method, such as an ELISA, to measure the amount of any one virus protein, or purify the virus away from other proteins.

The sample is purified virus with no other proteins. Can I use the UV absorbance at 280nm to measure the protein cocentration? The problem is I don't know what it the Molar absorption coefficient to substitute it in Beer Lambert's Law. Can you help with that ? 

Please, determine the protein concentration by HPLC-Soap-SEC method (please see file; prot deter Summary and HPLC-Soap-SEC protein determination method), and then analyse by using PDMD (protein-direct-microsequencing-deciphering) method (please see file; HepG2 fucoidan). 

Please use fresh specimens or samples stored at -196 C or -80 C in order to covalently bind proteins onto glass fibers effectively.   Used samples should be abandoned.   Therefore, prepared fresh sample specimens should be subdivided into Eppendorf tubes (reliable pack is possible to achieve), and stored.

Although I have not yet purified HAdV, I have searched for HAdV in my specimens by PDMD method.   HAdV is not present in fetal liver Hc cells and human breast milk, and HAdV-5 is not present in all my specimens.

Serum of 1y girl (before biotin therapy; state of biotin deficiency) has Maturation protein (Human adenovirus 6; HAdV-6)  at 7.7 μg/mg of cell protein.   HAdV is present at 8% (1/13) of Japanese serum.    

Cells of hepatoma (15y, boy, American Caucasoid) has Packaging protein 3 (HAdV-2; Human adenovirus 2) at 0.2 μg/mg of cell protein, and the healed HepG2 (with fucoidan for three days) has DNA polymerase (Human adenovirus type 40; HAdV-40) at 0.9 μg/mg of cell protein and Fiber protein (Protein IV; Human adenovirus type 12; HAdV-12) at 0.4 μg/mg of cell protein.   Although +ssRNA virus has disappeared in three days of fucoidan treatment at 0.102 mg/mL in vitro.   Hayashi K, et al. has reported that herpes simplex virus (HHV-1 and HHV-2; dsDNA virus) has been disappeared at 3 weeks of fucoidan feeding in mice (Defensive effects of a fucoidan from brown alga Undaria pinnatifida against herpes simplex virus infection. Int Immunopharmacol 8: 109-116, 2008).   Then, dsDNA virus of HAdV would also be disappeared by fucoidan for three weeks of therapy. 

Interestingly, all the five Japanese-liver specimens have HAdV, although no link to HCC is present. 

Liver of pseudo-liver cancer has E1A protein (HAdV-7) at 0.11, Hexon-associated protein (HAdV-7) at 0.29, Minor core protein (HAdV-2) at 1.6 μg/mg of tissue protein, respectively.

LC tissue (with leprosy, female) has Pre-protein VI　（Protein VI; HAdV-2) at 2.9 μg/mg of tissue protein.

HCC tissue (with PBC, female) has Early E1A 6 KD protein (HAdV-2) at 0.08 μg/mg of tissue protein.

LC tissue (No.6) has Probable early E4 11KD protein (HAdV-2) at 1.5, Early E3A 12.5 KD protein (HAdV-2) at 0.5, and Hypothetical protein ORF1 (Avian adenovirus 1; AAdV) at 0.6 μg/mg of tissue protein, respectively.

HCC tissue (No.6) has E3B 14.7 KD protein (HAdV-12) at 4.2, Fiber protein (HAdV-12) at 0.31, and 11.6 kD protein (HAdV-2) at 0.06  μg/mg of tissue protein, respectively.

The extinction coefficient of an individual protein can be calculated for a fully unfolded protein based on the amino acid sequence. In this case, however, there are 2 reasons why this won't work: the virus consists of several different proteins, and it contains nucleic acid, which also absorbs in the UV.

Since you have a purified virus, I think you should use a colorimetric or fluorometric protein assay. You will have to use a standard protein to calibrate the assay, and the choice of standard will affect the result. People usually use either bovine serum albumin or immunoglobulin G as the standard.

Unfortunately, these are destructive methods. If your sample is very precious and there isn't much of it, a fluorometric protein assay kit can be used, since such an assay is more sensitive than a colorimetric assay. By doing the assay in microplates, the sample volume required can be minimized. You would need a fluorescence plate reader.

Thanks so much for your answers. I think a BCA assay would be effective. Fortunately, it's ok if I lost some finding my protein concentration, I can still amplify some more.