Banca de QUALIFICAÇÃO: PATRICK CESAR ALVES TERREMATTE

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
DISCENTE : PATRICK CESAR ALVES TERREMATTE
DATA : 27/02/2019
HORA: 09:00
LOCAL: Núcleo de Pesquisas em Inovação em Tecnolgia da Informação - NPITI
TÍTULO:

Classification and Regression of clinical and genomic data of prostate cancer data through ensembles

PALAVRAS-CHAVES:

Prostate cancer. Genomics. Ensembles learning. Classification.

PÁGINAS: 48
GRANDE ÁREA: Ciências Exatas e da Terra
ÁREA: Ciência da Computação
RESUMO:

The analysis of clinical and genomic integrated data allows to classify the subtypes of cancer and provides an accurate prognosis. In this work, we analyze prostate cancer data from TCGA with correlation analysis and dimensionality reduction (PCA, t-SNE, and MDS).  The prostate cancer molecular subtypes (ERG, ETV1 / 4, FLI1, SPOP, FOXA1, IDH1, and others) were compared through a Kaplan-Meier and cumulative incidence curves for disease recovery and progression events. We did the regression for recovery estimation and classification for subtypes of prostate cancer, using ensembles learning strategies.  We apply the following regressors: kernel Support Vector Machine (SVM), Random Forrest, Bayesian and gradient boosted machines (GBM).  To classify, we used Kohonen's self-organizing maps (SOM), SVM and LogitBoost. Finally, we perform a differential expression analysis with R-Peridot with clusters of the ERG molecular subtype of prostate cancer.

MEMBROS DA BANCA:
Presidente - 347628 - ADRIAO DUARTE DORIA NETO
Interno - 1669545 - DANIEL SABINO AMORIM DE ARAUJO
Externa ao Programa - 1365498 - BEATRIZ STRANSKY FERREIRA