Submitted by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-06-29T13:54:15Z No. of bitstreams: 1 DIS_MAURICIO_BOFF_DE_AVILA_COMPLETO.pdf: 2632464 bytes, checksum: 1a170d33bfad88ea8a7aba65cb0f4bea (MD5) / Approved for entry into archive by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-06-29T13:54:25Z (GMT) No. of bitstreams: 1 DIS_MAURICIO_BOFF_DE_AVILA_COMPLETO.pdf: 2632464 bytes, checksum: 1a170d33bfad88ea8a7aba65cb0f4bea (MD5) / Made available in DSpace on 2017-06-29T13:54:34Z (GMT). No. of bitstreams: 1 DIS_MAURICIO_BOFF_DE_AVILA_COMPLETO.pdf: 2632464 bytes, checksum: 1a170d33bfad88ea8a7aba65cb0f4bea (MD5) Previous issue date: 2017-03-20 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior - CAPES / Microorganisms, in general, are the major agents of disease in humans. Data from the Brazilian Ministry of Health show bacterial diseases as the main causes of death in the country. In the therapy of these organisms, antibiotics are considered the most successful chemotherapy methods of 21st century medicine, as they represent the first, and only, line of combat against bacterial diseases. The development of new antibiotic drugs is becoming increasingly necessary, as bacterial resistance rates become higher each year. At this point, shikimate pathway is attractive to this type of research, since it is considered an essential pathway for the maintenance of these organisms in the environment, besides being absent in animals. The pathway is responsible for the formation of chorismate, precursor of aromatic amino acids (Phe, Trp and Tyr), folic acid and ubiquinones in the groups of organisms that presents it. The third reaction of the shikimate biosynthetic pathway is performed by the enzyme DHQD. In this step the reversible dehydration of the DHQ molecule is performed aiming to transform into 3-dehydroshikimate, the focus reaction of this study. In the search for new DHQD inhibitors, docking simulations were performed against the three-dimensional structure of a target protein, since it is a process that seeks to find, among possible orientations/conformations of a ligand in the active site, the one that presents the lower binding energy and, consequently, greater affinity. In addition to the docking simulations, machine learning methods were used to formulate polynomial scoring functions, based on the MVD scoring functions, which were able to predict protein/binder affinity. At the end of all the simulations and tests carried out throughout the project, we conclude that the Polscore56 equation was the most skilled to predict the affinity between the active site of DHQD and tested compounds. For this polynomial, the results of test set (? = 0,900; p-value = 0,037), AUC (74,686%), EF1 (540) and EF2 (159,23) were, in most of the categories evaluated, the best, confirming the formulated hypotheses on the equation and indicating it for further studies with the enzyme. / Microrganismos, em geral, apresentam-se como os principais agentes de doen?as em seres humanos. Dados do Minist?rio da Sa?de Brasileiro demonstram doen?as bacterianas como as principais causas de morte no pa?s. Na terap?utica desses organismos, os antibi?ticos s?o considerados os m?todos quimioter?picos de maior sucesso da medicina do s?culo XXI, pois representam a primeira, e ?nica, linha de combate contra doen?as bacterianas. O desenvolvimento de novas drogas antibi?ticas torna-se cada vez mais necess?rio, uma vez que os ?ndices de resist?ncia bacteriana se tornam mais altos a cada ano. Nesse ponto, a rota metab?lica do chiquimato ? atraente a esse tipo de pesquisa, por ser considerada uma via essencial para a manuten??o desses organismos no ambiente, al?m de estar ausente em animais. A via ? respons?vel pela forma??o do corismato, precursor de amino?cidos arom?ticos (Phe, Trp e Tyr), ?cido f?lico e ubiquinonas nos grupos de seres vivos onde est? presente. A terceira rea??o da via biossint?tica do chiquimato ? realizada pela enzima DHQD. Nesse passo ? realizada a desidrata??o revers?vel da mol?cula DHQ visando transform?-la em 3-desidrochiquimato, rea??o foco desse estudo. Na busca por novos inibidores de DHQD foram realizadas simula??es de docking de pequenos ligantes contra a estrutura tridimensional de uma prote?na alvo, pois ? um processo onde se visa encontrar, entre as poss?veis orienta??es/conforma??es de um ligante no s?tio ativo, aquela que apresenta a menor energia de liga??o e, consequente, maior finidade. Al?m das simula??es de docking, foram realizados m?todos de Aprendizagem de M?quina na formula??o de fun??es escores polinomiais, a partir de fun??es escores presentes no MVD, que fossem capazes de prever a afinidade entre prote?na/ligante. Ao final de todas as simula??es e testes realizados ao longo do projeto, chegamos ? conclus?o de que a equa??o Polscore56 apresentou-se como a mais h?bil para prever a afinidade entre o s?tio ativo de DHQD com compostos testados. Para esse polin?mio os resultados de test set (? = 0,900; p-value = 0,037), AUC (74,686%), EF1 (540) e EF2 (159,23) foram, na maioria das categorias avaliadas, os melhores, confirmando as hip?teses formuladas sobre a equa??o e indicando-a para estudos posteriores com a enzima.