Construção de uma plataforma para determinação in silico da alergenicidade e antigenicidade de proteínas

Abstract

With the advancement in sequencing of different genomes, the construction of genetically modified organisms (GMO), especially crop plants, has increased. One of the concerns involving the expression of different heterologous proteins is the possibility of development of allergic or antigenic reactions on final consumers. Thus, the WHO (World Health Organization) and the FAO (Food And Agriculture Organization Of The United Nations) proposed in silico tests to determine levels of antigenicity and allergenicity of proteins that could be used in the construction of GMO. Aiming to congregate several prediction algorithms that allow the in silico analysis of proteins and define their allergenic or antigenic potential, a platform called Platform for Analysis of Allergenicity and Antigenicity of Protein (PA³P) was built. In this platform the databases PIR, SwissProt/TrEMBL and GenBank and their AlgPred and Allermatch suites were congregated. Firstly, the identification of programming languages for each database was conducted with subsequent search through the sites aiming to identify their interfaces and how queries are performed. Some sites presented interface using REST, while others don't provide any information. Once the interfaces were identified, alternatives in the programming language Hypertext Preprocessor (PHP version 5.4.3) were sought. Such language was chosen because it has many libraries, a community of active development, adequate documentation and to be open acess. The WampServer (version 2.2) was used since it allows creating web applications with Apache2 (version 2.2.22), PHP and a MySQL database (version 5.5.24). For the sites with REST the set of operations defined by HTTP was used. For those that offered no interface, the most adequate operation was performing the remote sending of enquiry forms via cURL, respecting the original settings of the sites. For the interfaces of input and output data HyperText Markup Language (HTML) was used. The outputs were treated via regular expressions to get only the desired results, once some sites offer different answer formats to the WebService. To validate PA³P a sequence database was constructed, which contains 30 knowingly allergenic protein sequences, 30 non allergenic protein sequences, and 30 sequences from proteins whose levels of allergenicity and antigenicity were unknown. The sequences were submitted for analysis on both PA³P and sites individually. The validation process showed that our data are consistent with the data generated when the sequences were subjected to the algorithms individually, but with reduced analysis time. Thus, by using the referred platform, the user needs considerably less time when performing all necessary tests. The use of PA³P is considered reliable as indicated by the validation of this tool. Future works includes data persistence, the inclusion of statistical analysis, the addition of new algorithms of analysis, and graphical interface.