Please use this identifier to cite or link to this item:
http://hdl.handle.net/10609/98986
Title: | Empleo de técnicas de Machine Learning para la predicción de propiedades ADME-Tox: Toxicidad |
Author: | Vela Castro, Alberto |
Director: | Enciso Carrasco, Marta |
Tutor: | Cánovas Izquierdo, Javier Luis |
Keywords: | machine learning support vector machine artificial neural networks toxicity database statistical analysis |
Issue Date: | 4-Jun-2019 |
Publisher: | Universitat Oberta de Catalunya (UOC) |
Abstract: | There is a growing predilection for the application of in silico techniques in the development and discovery of new drugs opposite of costly and laborious laboratory techniques. These are machine learning techniques. The Master Final Project (TFM) consisted in the analysis of the best current techniques of machine learning for the prediction of ADME-Tox property, toxicity. Once a theoretical comparison has been made, a practice will be carried out with a real database where it will be possible to observe the different efficiencies in the prediction of the different proposed models. The methodology was carried out with the free software of R and the "rcdk" package for the generation of the descriptors, followed by a pre-processing of the data and a subsequent generation of the algorithms with their proper comparison. The algorithm that was differentiated from the rest by its characteristics was the Decision Tree with an accuracy of 0.88 and a kappa index of 0.72 for this type of data. Thanks to the fact that it is capable of operating with low data volumes, few levels and above all because of its ability to exclude unimportant features. It could be concluded that for databases with a large number of numerical descriptors and few values, the ideal algorithm would be the Decision Tree. |
Language: | Spanish |
URI: | http://hdl.handle.net/10609/98986 |
Appears in Collections: | Bachelor thesis, research projects, etc. |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
velacastroaTFM0619memoria.pdf | Memoria del TFM | 925,65 kB | Adobe PDF | ![]() View/Open |
Share:


This item is licensed under a Creative Commons License