QGRS-Conserve: a computational method for discovering evolutionarily conserved G-quadruplex motifs

Abstract

Background: Nucleic acids containing guanine tracts can form quadruplex structures via non-Watson-Crick base pairing. Formation of G-quadruplexes is associated with the regulation of important biological functions such as transcription, genetic instability, DNA repair, DNA replication, epigenetic mechanisms, regulation of translation, and alternative splicing. G-quadruplexes play important roles in human diseases and are being considered as targets for a variety of therapies. Identification of functional G-quadruplexes and the study of their overall distribution in genomes and transcriptomes is an important pursuit. Traditional computational methods map sequence motifs capable of forming G-quadruplexes but have difficulty in distinguishing motifs that occur by chance from ones which fold into G-quadruplexes.

Results: We present Quadruplex forming 'G'-rich sequences (QGRS)-Conserve, a computational method for calculating motif conservation across exomes and supports filtering to provide researchers with more precise methods of studying G-quadruplex distribution patterns. Our method quantitatively evaluates conservation between quadruplexes found in homologous nucleotide sequences based on several motif structural characteristics. QGRS-Conserve also efficiently manages overlapping G-quadruplex sequences such that the resulting datasets can be analyzed effectively.

Conclusions: We have applied QGRS-Conserve to identify a large number of G-quadruplex motifs in the human exome conserved across several mammalian and non-mammalian species. We have successfully identified multiple homologs of many previously published G-quadruplexes that play post-transcriptional regulatory roles in human genes. Preliminary large-scale analysis identified many homologous G-quadruplexes in the 5'- and 3'-untranslated regions of mammalian species. An expectedly smaller set of G-quadruplex motifs was found to be conserved across larger phylogenetic distances. QGRS-Conserve provides means to build datasets that can be filtered and categorized in a variety of biological dimensions for more targeted studies in order to better understand the roles that G-quadruplexes play.

Figure 1

G-quadruplex structure. Left: an intramolecular G-quadruplex formed by a G₃AG₃CUG₃CG₃ RNA motif. Right: a G-tetrad structure.

Figure 2

Overlap percentage computation. The extension of sequence length using 50% of the total sequence length as padding inflates overlap scores for proximate sequences.

Figure 3

Algorithm stages for SNCA mRNA. Overview of the algorithm stages for QGRS identification (independently) on each mRNA, performing a semi-global alignment on the homologs, and finally evaluating the QGRS pair for conservation.

Figure 4

Overlapping QGRS yielding different G-scores. A small sample of hundreds of overlapping QGRS found in the ‘GGATGAGCCTGTGGGTGGAGGGCAGTGGAGCGGGCTGGG’ sequence of human GREB1 gene.

Figure 5

Choosing QGRS considered for conservation calculations. The number next to each QGRS motif represents its calculated G-score. This figure demonstrates how it is possible that QGRS with highest G-score do not necessarily yield the highest conservation score.