Please use this identifier to cite or link to this item:
Title: Benchmark de algoritmos para la detección de CNVs con resolución de exón a partir de datos NGS de panel
Author: Moreno Cabrera, José Marcos
Director: Gel Moreno, Bernat
Tutor: Ventura Royo, Carles  
Others: Universitat Oberta de Catalunya
Keywords: CNV
next generation sequencing
Issue Date: Jan-2018
Publisher: Universitat Oberta de Catalunya
Abstract: Next Generation Sequencing (NGS) is a key technology for detecting small variants in the genetic diagnostics of hereditary diseases. However, detection of larger variants as copy number variants (CNVs) from NGS data remains a challenge. The most used technique for CNV detection is multiplex ligation-dependent probe amplification (MLPA), which implies important costs and time. Finding a screening technique would allow us to decrease the number of necessary MLPAs, and therefore, resources may be saved. Most CNV calling algorithms perform well when calling large CNVs (in the order of megabases) but are not able to reliably detect small CNVs affecting only one or few exons. In addition, most of them are designed to work with whole exome or whole genome data and have problems with the more sparsed data generated by NGS panels. In this work, it has been perfomed a benchmark of CNV calling algortihms which have shown to perform well with NGS data at single exon resolution: DECoN, CoNVaDING and panelcn.MOPS. These algorithms have been evaluated over a publicy available and validated dataset of 96 samples. Moreover, each algorithm has been optimized using an optimization algorithm to improve sensitivity and specificity offered by default parameter values. Results revealed panelcn.MOPS reached sufficient sensitivity in the dataset to be used for CNV screening prior to MLPA validation.
Language: Spanish
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
jpuntomarcosTFM0118Memoria.pdfMemoria del TFG936 kBAdobe PDFView/Open

This item is licensed under a Creative Commons License Creative Commons