Relaxed formation rules for G-quadruplex prediction

Abstract

As a secondary structure of DNA, G-quadruplexes are formed by folding of genomic DNA so that four repetitive guanine tracts are brought together. Some of these structures are shown to have significant role on gene regulation, however, detecting G-quadruplex formation by genomic DNA presents a significant obstacle as the these structures are easily affected by harsh experimental conditions. Instead, it is shown that prediction of these secondary structures is possible through analysis of the DNA sequence. However, the most commonly used algorithm was shown to miss two types of imperfect G-quadruplexes (i) with bulges or mismatches in the guanine-repeats, and also ones (ii) with an extraordinarily long loop. Here, we present a tool for the detection of these structures including such rudimentary but possible G-quadruplexes using G4Catchall, improving the overall prediction efficiency. © 2018 IEEE.