O uso de redes neurais artificiais na análise de dados de câncer de pulmão

• Ano de defesa: 2022
• Autor(a) principal: Lima, Débora Virgínia da Costa e
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Rio Grande do Norte; Brasil; UFRN; PROGRAMA DE PÓS-GRADUAÇÃO EM BIOINFORMÁTICA
• Palavras-chave em Português: Câncer de pulmão; Aprendizado profundo; Dados genômicos; Assinatura gênica; Sobrevida
• Link de acesso: https://repositorio.ufrn.br/handle/123456789/48470

Lung cancer represents the leading cause of cancer death worldwide and has a high incidence. Like other types of cancer, it can occur due to different causes, from genetics to environmental ones, so studies carried out using different types of data may be relevant for the control of this neoplasm, especially when considering factors that have an impact on patient survival. In the context of lung cancer, this study uses deep learning to predict patient survival. Clinical and molecular data from TCGA (The Cancer Genome Atlas) databases were obtained for the LUSC (Lung Squamous Cell Carcinoma) and LUAD (Lung Adenocarcinoma) cohorts, followed by the analysis of the genomic alterations, and application of neural networks using as input the frequently mutated genes for each cohort, selection of key genes and validation with another database. The cohorts showed differences in survival among themselves when subjected to the Kaplan-Meier method and the Log-Rank test. In the genomic analysis, all genes with a mutation frequency above 15% were selected, and 34 genes were found for LUAD and 32 for LUSC. The use of these genes as input in the constructed networks made it possible to generate the LUSC and LUAD networks with 100% accuracy, identifying, according to the mutations, the vital status of the patient. In addition, a LUSC network was also obtained using another LUSC-KR database as validation, which reached 99% accuracy. In this way, this work showed that the use of genes with frequent mutations associated with deep learning is a robust tool and allows predicting the survival of patients with lung cancer.