O câncer cervical afeta milhões de mulheres em todo o mundo a cada ano. A maioria dos casos de câncer cervical é causada pelo vírus do papiloma humano (HPV) que é sexualmente transmissível. Cerca de 40 tipos de HPV infectam o colo do útero e estes são designados como sendo de alto ou de baixo risco com base no seu potencial para provocar lesões de alto grau e câncer. A oncoproteína E7 do HPV está diretamente envolvida no aparecimento de câncer de colo do útero. Esta se associada com a proteína pRb e outros alvos celulares que promovem a imortalização celular e carcinogênese. Apesar de muito progresso nos estudos sobre os HPVs de alto risco, ainda não existe uma terapêutica adequada para o tratamento das lesões e câncer causados por este vírus. Este trabalho teve como objetivo entender as diferenças estruturais entre E7 de alto e baixo risco e sugerir, através de análises de bioinformática, possíveis sítios de ligação e inibidores para a E7. Esta é a primeira descrição da modelagem e análise de dinâmica molecular de quatro estruturas tridimensionais completas da E7 dos tipos de alto risco (HPV tipos 16 e 18), de baixo risco (HPV tipo 11) e não relacionadas ao câncer cervical (HPV tipo 1A). Os modelos foram construídos por uma abordagem híbrida usando modelagem por homologia e ab initio. Os modelos foram usados em simulações de dinâmica molecular por 50 ns, sob condições normais de temperatura e pressão. A desordem intrínseca da sequência da proteína E7 foi avaliada com o uso de ferramentas in silico. Os domínios N-terminal de todas as E7 estudadas, mesmo as de alto risco, exibiram estruturas secundárias depois da modelagem. Nas análises da trajetória da dinâmica molecular, as E7s dos HPVs dos tipos 16 e 18 apresentaram maior instabilidade nos seus domínios N-terminais em relação aos do HPV dos tipos 11 e 01. No entanto, esta variação não afetou a conformação das estruturas secundárias durante a simulação. A análise com ANCHOR indicou que as regiões CR1 e CR2 regiões dos tipos de HPV 16 e 18 contêm possíveis alvos para a descoberta da droga. Já a região CR3 do domínio C-terminal indicou estabilidade nas análises in silico e, por isso, foi usada como alvo de busca de modelos farmacofóricos e docking macromolecular. A proteína usada como modelo foi a E7 do HPV tipo 45 resultante de análises de ressonância magnética nuclear (RMN) e depositada no banco de dados de proteína (ID: 2F8B). Foram selecionados por análises sequenciais de busca farmacofórica, docking e re-docking, 19 compostos (extraídos de amplas bibliotecas de pequenos ligantes) com potencial para candidatos a inibidores da E7. Eles foram avaliados quanto a sua função de pontuação, mapas de interação receptor-ligante e toxicidade e os melhores foram indicados para estudos futuros. / Cervical cancer affects millions of women around the world each year. Most cases of cervical cancer are caused by human papilloma virus (HPV) which is sexually transmitted. About 40 types of HPV infect the cervix and these are designated as being at high or low risk based on their potential to cause high-grade lesions and cancer. The E7 oncoprotein from HPV is directly involved in the onset of cervical cancer. It associates with the pRb protein and other cellular targets that promote cell immortalization and carcinogenesis. Although the progress in studies with high-risk HPVs there is still no adequate therapy for the treatment of lesions and cancers caused by this virus. This study aimed to understand the structural differences between E7 of high and low risk and suggest, with the aid of bioinformatics analyzes, possible binding sites and inhibitors for the E7. This is the first description of the modeling and molecular dynamics analysis of four complete three-dimensional structures of E7 from high-risk types (HPV types 16 and 18), low risk (HPV type 11) and that not related to cervical cancer (HPV 01). The models were constructed by a hybrid approach using homology modeling and ab initio. The models were used in molecular dynamics simulations for 50 ns, under normal temperature and pressure. The intrinsic disorder of the E7 protein sequence was assessed using in silico tools. The N-terminal domains of all E7s, even the high-risks, showed secondary structures after modeling. In the trajectory analyzes of molecular dynamics, the E7s of HPV types 16 and 18 showed high instability in their N-terminal domains than those of HPV types 11 and 01, however, this variation did not affect the conformation of secondary structures during the simulation. The analysis with ANCHOR indicated that regions CR1 and CR2 regions of types of HPV 16 and 18 contain possible targets for drug discovery. The CR3 region of the C-terminal domain indicated stability by in silico analyzes and was therefore used as target to search for pharmacophoric models and \"docking\". The protein used as a model was the E7, from HPV type 45, constructed by analysis of nuclear magnetic resonance (NMR) and deposited in the protein data bank (ID: 2F8B). It was selected 19 compounds as potential candidates for E7 inhibitors (extracted from large libraries of small ligands) using sequential pharmacophore search, docking and re-docking analyzes. They were evaluated for their scoring function, maps of receptor-ligand interactions and toxicity and the best suited were indicated for future studies.

A great challenge in bioinformatics is to accurately predict protein structure and function from its amino acid sequence, including annotation of protein domains, identification of protein disordered regions and detecting protein stability changes resulting from amino acid mutations. The combination of bioinformatics, genomics and proteomics becomes essential for the investigation of biological, cellular and molecular aspects of disease, and therefore can greatly contribute to the understanding of protein structures and facilitating drug discovery. In this thesis, a PREDICTOR, which consists of three machine learning methods applied to three different but related structure bioinformatics tasks, is presented: using profile Hidden Markov Models (HMMs) to identify remote sequence homologues, on the basis of protein domains; predicting order and disorder in proteins using Conditional Random Fields (CRFs); applying Support Vector Machines (SVMs) to detect protein stability changes due to single mutation. To facilitate structural instability and disease studies, these methods are implemented in three web servers: FISH, OnD-CRF and ProSMS, respectively. For FISH, most of the work presented in the thesis focuses on the design and construction of the web-server. The server is based on a collection of structure-anchored hidden Markov models (saHMM), which are used to identify structural similarity on the protein domain level. For the order and disorder prediction server, OnD-CRF, I implemented two schemes to alleviate the imbalance problem between ordered and disordered amino acids in the training dataset. One uses pruning of the protein sequence in order to obtain a balanced training dataset. The other tries to find the optimal p-value cut-off for discriminating between ordered and disordered amino acids.  Both these schemes enhance the sensitivity of detecting disordered amino acids in proteins. In addition, the output from the OnD-CRF web server can also be used to identify flexible regions, as well as predicting the effect of mutations on protein stability. For ProSMS, we propose, after careful evaluation with different methods, a clustered by homology and a non-clustered model for a three-state classification of protein stability changes due to single amino acid mutations. Results for the non-clustered model reveal that the sequence-only based prediction accuracy is comparable to the accuracy based on protein 3D structure information. In the case of the clustered model, however, the prediction accuracy is significantly improved when protein tertiary structure information, in form of local environmental conditions, is included. Comparing the prediction accuracies for the two models indicates that the prediction of mutation stability of proteins that are not homologous is still a challenging task. Benchmarking results show that, as stand-alone programs, these predictors can be comparable or superior to previously established predictors. Combined into a program package, these mutually complementary predictors will facilitate the understanding of structural instability and disease from protein sequence.

O câncer cervical afeta milhões de mulheres em todo o mundo a cada ano. A maioria dos casos de câncer cervical é causada pelo vírus do papiloma humano (HPV) que é sexualmente transmissível. Cerca de 40 tipos de HPV infectam o colo do útero e estes são designados como sendo de alto ou de baixo risco com base no seu potencial para provocar lesões de alto grau e câncer. A oncoproteína E7 do HPV está diretamente envolvida no aparecimento de câncer de colo do útero. Esta se associada com a proteína pRb e outros alvos celulares que promovem a imortalização celular e carcinogênese. Apesar de muito progresso nos estudos sobre os HPVs de alto risco, ainda não existe uma terapêutica adequada para o tratamento das lesões e câncer causados por este vírus. Este trabalho teve como objetivo entender as diferenças estruturais entre E7 de alto e baixo risco e sugerir, através de análises de bioinformática, possíveis sítios de ligação e inibidores para a E7. Esta é a primeira descrição da modelagem e análise de dinâmica molecular de quatro estruturas tridimensionais completas da E7 dos tipos de alto risco (HPV tipos 16 e 18), de baixo risco (HPV tipo 11) e não relacionadas ao câncer cervical (HPV tipo 1A). Os modelos foram construídos por uma abordagem híbrida usando modelagem por homologia e ab initio. Os modelos foram usados em simulações de dinâmica molecular por 50 ns, sob condições normais de temperatura e pressão. A desordem intrínseca da sequência da proteína E7 foi avaliada com o uso de ferramentas in silico. Os domínios N-terminal de todas as E7 estudadas, mesmo as de alto risco, exibiram estruturas secundárias depois da modelagem. Nas análises da trajetória da dinâmica molecular, as E7s dos HPVs dos tipos 16 e 18 apresentaram maior instabilidade nos seus domínios N-terminais em relação aos do HPV dos tipos 11 e 01. No entanto, esta variação não afetou a conformação das estruturas secundárias durante a simulação. A análise com ANCHOR indicou que as regiões CR1 e CR2 regiões dos tipos de HPV 16 e 18 contêm possíveis alvos para a descoberta da droga. Já a região CR3 do domínio C-terminal indicou estabilidade nas análises in silico e, por isso, foi usada como alvo de busca de modelos farmacofóricos e docking macromolecular. A proteína usada como modelo foi a E7 do HPV tipo 45 resultante de análises de ressonância magnética nuclear (RMN) e depositada no banco de dados de proteína (ID: 2F8B). Foram selecionados por análises sequenciais de busca farmacofórica, docking e re-docking, 19 compostos (extraídos de amplas bibliotecas de pequenos ligantes) com potencial para candidatos a inibidores da E7. Eles foram avaliados quanto a sua função de pontuação, mapas de interação receptor-ligante e toxicidade e os melhores foram indicados para estudos futuros. / Cervical cancer affects millions of women around the world each year. Most cases of cervical cancer are caused by human papilloma virus (HPV) which is sexually transmitted. About 40 types of HPV infect the cervix and these are designated as being at high or low risk based on their potential to cause high-grade lesions and cancer. The E7 oncoprotein from HPV is directly involved in the onset of cervical cancer. It associates with the pRb protein and other cellular targets that promote cell immortalization and carcinogenesis. Although the progress in studies with high-risk HPVs there is still no adequate therapy for the treatment of lesions and cancers caused by this virus. This study aimed to understand the structural differences between E7 of high and low risk and suggest, with the aid of bioinformatics analyzes, possible binding sites and inhibitors for the E7. This is the first description of the modeling and molecular dynamics analysis of four complete three-dimensional structures of E7 from high-risk types (HPV types 16 and 18), low risk (HPV type 11) and that not related to cervical cancer (HPV 01). The models were constructed by a hybrid approach using homology modeling and ab initio. The models were used in molecular dynamics simulations for 50 ns, under normal temperature and pressure. The intrinsic disorder of the E7 protein sequence was assessed using in silico tools. The N-terminal domains of all E7s, even the high-risks, showed secondary structures after modeling. In the trajectory analyzes of molecular dynamics, the E7s of HPV types 16 and 18 showed high instability in their N-terminal domains than those of HPV types 11 and 01, however, this variation did not affect the conformation of secondary structures during the simulation. The analysis with ANCHOR indicated that regions CR1 and CR2 regions of types of HPV 16 and 18 contain possible targets for drug discovery. The CR3 region of the C-terminal domain indicated stability by in silico analyzes and was therefore used as target to search for pharmacophoric models and \"docking\". The protein used as a model was the E7, from HPV type 45, constructed by analysis of nuclear magnetic resonance (NMR) and deposited in the protein data bank (ID: 2F8B). It was selected 19 compounds as potential candidates for E7 inhibitors (extracted from large libraries of small ligands) using sequential pharmacophore search, docking and re-docking analyzes. They were evaluated for their scoring function, maps of receptor-ligand interactions and toxicity and the best suited were indicated for future studies.