A Web Based Method for Computing Endpoint Titer and Concentration of Antibody/Antigen

**G. P. S. Raghava^* and Javed N. Agrewala**

Indraprastha Institute of Information Technology

New Delhi, India

^*Address for Correspondance

Dr. G. P. S. raghava, Professor

Department of Computational Biology

Indraprastha Institute of Information Technology

Okhla Phase 3, New Delhi

INDIA

Email: raghava@imtech.ernet.in

Web: http://imtech.ernet.in/raghava

Phone:+91-172-690908

Fax: +91-172-690632

Abstract

In this report we have described a web-server for calculating the endpoint titers and concentrations of antibody/antigen (Ab/Ag) from the optical density (OD) taken from ELISA data. The server utilize a graphical method (Raghava et al., 1992) for determining the concentration of either the antibody or the antigen. In order to calculate the endpoint titer, first we fit the OD verses concentration of control data using a least square curve-fitting method. Then we fit the OD verses concentration of standard sample using graphical method. Finally, we determine the intersection or nearest point of two curves which we have called the endpoint titer. In order to calculate concentrations of the antibody/antigen of unknown samples, we have to first fit OD verses concentrations of the known samples using graphical method and to determine the linear interpolation and hyperbolic formulas. Then we calculate the concentrations of the unknown samples from their OD using these formula's. This web-server utilizes the CGI program written in PERL and java-script, which make server live and interactive (HTTP://imtech.ernet.in/raghava/abag/).

Description

The endpoint titer is routinely used in immunology to measure the secretion of antibody. In order to compute the titer of an antibody, Peterfy et al. (1983) used the low endpoint titer (10% of maximum OD). Caulfield and Shaffer (1984) develop a program and calculated the endpoint titer using OD as 0.1. They fitted the standard curve using local method. Recently, Tremain (1993) developed a program for calculating endpoint titer of antibody from ELISA data. The standard curve was fitted using iterative simplex algorithm (Nelder and Meed, 1965; Tijssen, 1885). This method allow the user to select the cut-off point for calculating endpoint titer. However, 10% of the maximum OD was recommended in this method.

In these methods, authors uses the different ODs for endpoint titer. None of them has taken into consideration the effect of antibody concentration on the OD in absence of any interaction (control data). The background OD varies with the variation in the concentration of the antibody, as well as different antibodies give different background. In order to consider the effect of concentration of antibody on background OD and other factors, a new method has been developed to compute the endpoint titer of the antibody.

The method described in this report utilize the OD verses(vs) antibody concentration of known samples using graphical method which is more accurate and sensitive (Raghava et al., 1992). The graphical method combines the power of local and global fitting method. The OD vs log concentration of antibody of the control data was fitted using least square curve fitting method. We then calculated the intersection of the standard curve and the control curve or nearest point of the two curves. This allows the method to incorporate the effect of background OD due to non specific noise induced by the antibody.

In the past, number of computer programs have been developed for calculating the Ag/Ab concentrations (Slezak et al., 1983; Caulfield et al., 1984; Studnicka, 1987; Studnicka, 1991). In ELISA procedure an equation is derived using standards to measure Ab/Ag concentration of unknown samples. This is done by a series of dilutions of known standards to derive an equation by fitting a standard curve. Which serve as an internal calibration for the unknown samples on the plate. The equation of standard is used to measure the Ab/Ag concentration of unknown samples.

Previously, we have developed a computer program called ELISAeq (Raghava et al., 1992) which was designed to determine the concentration of Ab/Ag using ELISA data. In ELISAeq, the graphical method was used which utilizes both the linear regression and hyperbolic regression methods for calculating Ab/Ag concentration (Raghava et al., 1992; Raghava et al., 1994). The linear regression method used in this program works only in the semilogarithmic linear(Sl)-range but it is more sensitive than the hyperbolic regression method. We have also earlier shown that our graphical method is more sensitive than the previously published methods (Raghava et al., 1992).

In order to provide the service world wide, we have developed a web-server which allows to compute the endpoint titer and concentration of antibody/antigen from ELISA data. In case the concentration is not known, then it allow, to compute endpoint titer and quantification of antibody in terms of dilution factor.

Hardware/ Software requirement

The user who have access to Internet and web browser can use this web server (HTTP://www.imtech.ernet.in/raghava/abag/). These web pages can be loaded on any computer which can run the web server and have the PERL language interpreter. The web pages use the java script and CGI script, written in PERL.

References

Caulfield, M.J. and Shaffer, D. (1984) A computer program for the evaluation of ELISA data obtained using an automated microtiter plate absorbaence reader. J. Immunol. Methods 14, 205.

Nelder, J. A. and Meed, R. (1965) A simplex method for function minimization, Computer Journal 7, 308.

Peterfy, F. P., Kuusela and Makela (1983) Affinity requirements for antibody assay mapped with monoclonal antibodies. J. Immunol. Methods, 138, 1809.

Raghava, G.P.S. and J.N. Agrewala. 1994. Method for determining the affinity of monoclonal antibody using non-competitive ELISA : A computer program. J. Immunoassay 15, 115.

Raghava, G. P. S., Joshi, A. K. and Agrewala, J. N. (1992) Calculation of antibody and antigen concentrations from ELISA data using a graphical method. J. Immunol. Methods, 153, 263.

Slezak, T.R., Vanderlaan, M. and Jensen, H. (1983) A computer-based data analysis system for enzyme-linked immunosorbent assays.

Studnicka, G.M. (1987) Hyperbolic regression analysis for kinetics, electrophoresis, ELISA, RIA, Bradford, Lowery, and other applications. Comput. Applic. Biosci. 3, 9.

Studnicka, G.M. (1991) ELISA assay optimization using hyperbolic regression. Comput. Applic. Biosci. 7, 217.

Tijssen,P. (1985) Practice and Theory of enzyme Immunoassays. Elsevier, Amsterdam, 385.