Please use this identifier to cite or link to this item: http://hdl.handle.net/10773/34283
Title: Analysis of single-strand exceptional word symmetry in the human genome: new measures
Author: Afreixo, Vera
Rodrigues, João M. O. S.
Bastos, Carlos A. C.
Keywords: Effect size measure
Exceptional symmetry
Hypothesis testing
Single-strand symmetry
Issue Date: Apr-2015
Publisher: Oxford University Press
Abstract: Some previous studies suggest the extension of Chargaff's second rule (the phenomenon of symmetry in a single DNA strand) to long DNA words. However, in random sequences generated under an independent symbol model where complementary nucleotides have equal occurrence probabilities, we expect the phenomenon of symmetry to hold for any word length. In this work, we develop new statistical methods to measure the exceptional symmetry. Exceptional symmetry is a refinement of Chargaff's second parity rule that highlights the words whose frequency of occurrence is similar to that of its reversed complement but dissimilar to the frequencies of occurrence of other words which contain the same number of nucleotides A or T. We analyze words of lengths up to 12 in the complete human genome and in each chromosome separately. We assess exceptional symmetry globally, by word group, and by word. We conclude that the global symmetry present in the human genome is clearly exceptional and significant. The chromosomes present distinct exceptional symmetry profiles. There are several exceptional word groups and exceptional words with a strong exceptional symmetry.
Peer review: yes
URI: http://hdl.handle.net/10773/34283
DOI: 10.1093/biostatistics/kxu041
ISSN: 1465-4644
Publisher Version: https://doi.org/10.1093/biostatistics/kxu041
Appears in Collections:CIDMA - Artigos
DETI - Artigos
DMat - Artigos
IEETA - Artigos

Files in This Item:
File Description SizeFormat 
kxu041.pdf491.99 kBAdobe PDFView/Open


FacebookTwitterLinkedIn
Formato BibTex MendeleyEndnote Degois 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.