Introduction Clostridium novyi causes gangrene in humans and animals. Among the various toxins produced by this pathogen (Hatheway, 1990), alpha toxin has a molecular weight of approximately 200-250 kDa and enzymatic activity on Rho receptors in mammalian cells, which cause vascular permeability and cell rounding in culture (Bette et al., 1991). Homology studies have shown that alpha toxin belongs to the large clostridial cytotoxin family (Hofmann et al., 1995). This family comprises glycosyltransferase by modifying small GTP-binding proteins, such as Rho and Ras, which inhibit signal pathways, thereby causing cytoskeletal disintegration (Oksche et al., 1992), which in turn causes cell death. Alpha toxin has a tripartite structure in which the enzymatic activity is related to a 551-amino acid on the N-terminal, known as the catalytic domain (Busch et al., 2000). The C-terminal of the toxin is the receptor-binding domain, and the hydrophobic domain in the central region is required for the translocation of the toxin into cytosol (von Eichel-Streiber et al., 1996). Vaccination is an effective means for the control of this disease. Whole-cell and toxoid vaccines are available, yet they have drawbacks such as incomplete inactivation and possible side effects. The use of recombinant DNA technology could result in antigens with high purity and safety (Fox and Klass, 1989). Protective antibodies against this recombinant antigen could react to the whole antigen and effective vaccines, producing the need to identify B-cell epitopes (Atassi et al., 2012). Therefore, epitope prediction is useful for the production of subunit and synthetic peptide vaccines and antibodies or antisera for competition assays. Identification of B-cell epitope locations has been carried out with both experimental and computational methods. In some experimental approaches, overlapping peptide was synthesized and bound to the antibody against the whole antigen (Atassi et al., 2011). Experimental techniques are time-consuming and expensive; furthermore, in the case of large proteins, it is laborious to carry out epitope mapping. Therefore, several prediction tools have been developed for identifying appropriate parts of protein as epitopes (Irving et al., 2001). This prediction is based on using computational methods and resources to extract immunological information and is known as immunoinformatics (Tomar and De, 2014). In this study, we designed a strategy to produce highly antigenic regions based on immunoinformatics methods in the recognized epitopic region, according to continuous, discontinuous, and T-cell epitopes. Production of antibodies was elicited by the immunization of mice