Please use this identifier to cite or link to this item:
Title: Predicció dels resultats d'edició genòmica amb CRISPR-Cas9 i base editors a partir de la seqüència de la regió modificada
Author: Expòsit Goy, Marc
Director: Prados Carrasco, Ferran
Tutor: Pla Planas, Albert
Keywords: CRISPR gene editing outcomes
classification models
machine learning
Issue Date: 24-Jun-2020
Publisher: Universitat Oberta de Catalunya (UOC)
Abstract: The potential use of gene editing technologies as therapeutics is limited by the lack of control in the outcomes of gene editing. These outcomes are determined, in part, by the sequence of the edited region. In this work, a machine learning model is used to predict the outcomes of CRISPR-Cas9 gene editing from the sequence of the gRNA. This model could be used to improve gRNA design so that gene editing outcomes are controlled. While previous studies introduce mutations in synthetic target sequences, in this work insertions are done in 1785 unique regions of the genome. Hence, experimental data reflect more closely the conditions in which the techniques would be applied in the clinic. Analyzing the target genomic regions reveals that sequencing coverage is not enough to quantify gene editing outcomes. Hence, these are simulated using previously developed models. Simulated data is treated in the same was as it would be done with experimental data. The gRNA efficiency prediction model is developed as a binary classifier, and logistic regression is the algorithm with the higher accuracy. The predictions are similar between this model and the original model used to simulate the data. The model to predict gene editing outcomes is planned using two different approaches that require further development. In brief, this work defines the steps and develops all the processes needed to go from experimental genomic data to the training of a computational model that predicts gene editing outcomes from the gRNA sequence.
Language: Catalan
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
mexpositTFM0620memòria.pdfMemòria del TFM2.27 MBAdobe PDFView/Open

This item is licensed under a Creative Commons License Creative Commons