Please use this identifier to cite or link to this item:
http://hdl.handle.net/10609/121346
Title: | Análisis de técnicas de clasificación de perfiles taxonómicos para predecir trastornos de la enfermedad inflamatoria intestinal |
Author: | Castillo Rosa, Eva |
Tutor: | Adsuar Gómez, Antonio Jesús |
Others: | Canovas Izquierdo, Javier Luis Maceira, Marc |
Abstract: | The inflammatory bowel disease comprises a wide range of disorders with similar symptoms. Therefore, studying the bacteria present in the microbiota of patients is key for the diagnosis and treatment of these diseases. A thorough study of different available classification algorithms is crucial to find the optimal ones and apply it to the discovery of biomarkers or, ultimately, clinical diagnosis. In this study, the microbial diversity of biopsy samples from healthy, Crohn's disease or ulcerative colitis patients was analysed with QIIME 2 software. Various supervised machine learning methods have been applied from bacterial relative abundance data to sample classification. Finally, an interactive web application has been developed in order to adapt the optimal models to the user's input data. Although some linear models show similar performance to complex ones, the model with the highest performance is random forest. Besides, choosing a good dimensionality reduction method is important when applying machine learning on microbiome data. Just as crucial as making these analyses available to the entire scientific community, so that large-scale studies can be done. |
Keywords: | microbiota machine learning shiny |
Document type: | info:eu-repo/semantics/masterThesis |
Issue Date: | Jun-2020 |
Publication license: | http://creativecommons.org/licenses/by-nc-nd/3.0/es/ |
Appears in Collections: | Trabajos finales de carrera, trabajos de investigación, etc. |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
ecastillorosTFM0620memoria.pdf | Memoria del TFM | 1,44 MB | Adobe PDF | View/Open |
Share:
This item is licensed under a Creative Commons License