Previsão de respostas a tratamentos de linhas celulares cancerígenas

Abstract

Cancer is one of the diseases with the highest mortality rate in the world. To understand the dif-ferent origins of the disease, and to facilitate the development of new ways to treat it, laboratoriescultivate,in vitro, cancer cells (cell lines), taken from patients with cancer. These cell lines enableresearchers to test new approaches and to have an appropriate procedure for comparison of results.At EMBL-EBI Institute (Cambridge, UK) an initial study was performed in which the effectof a large number of molecules was tested, in laboratory, in the treatment of cell lines with varioustypes of cancer. This study also included the use of Machine Learning algorithms, building modelsto predict the degree of efficacy of those drugs in cancer treatment.The methods used in the reported initial study were based on algorithms that construct "propositional-like" models. The results reported are promising but, we think, can be improved. Another lim-itation of the algorithms used in the original study is the absence or severe comprehensibilitylimitations of the models constructed. In areas of Life Sciences, the possibility of understandingthe forecast model is an asset to help the specialist to understand the phenomenon that producedthe data.Our thesis work has two main objectives: i) improve the performance of forecasting methods;and ii) understandability of the models constructed. To meet these objectives we proposed the useof Inductive Logic Programming (ILP) and Evolutionary Computation algorithms