Effects of T-Quadruplex on Affymetrix GeneChip® Data (A Data Analytical Approach

Abstract

The Karst Hoogsteen theory that bases can also pair up in many different ways opens up another door of research for scientists. It was observed in previous research that four runs of contiguous Guanines form unusual structures called G-Quadruplex structures and some abnormal behaviors are found on microarray data particularly on Affymetrix GeneChip® data. These behaviors were associated with the formation of G-quadruplex structures due to the presence of G-stack probes. These probes were not correlated with their other member probes while they were correlated with each other regardless of the genes/probe sets. Hence, G-stack probes were suggested unreliable for gene expression measurement. This left another question that if G-G can affect GeneChip® data then interaction among other bases like T-T, A-A and C-C may also be problematic for GeneChips®. The main objective of this research is to analyze the effects of thymine-thymine binding at probe level data of Affymetrix GeneChip®. Data of three different GeneChips® designed for organisms of three different kingdoms are downloaded from NCBI GEO repository. The data of Human (Kingdom Animalia), Arabidopsis Thaliana (Kingdom Plantae) and Escherichia Coli (Kingdom Bacteria) chips are used for analysis. Correlation among the T-stack probes was calculated to verify if they are behaving like G-stack probes or not. Our results suggest that thymine-thymine binding does not affect any of the chips taking into consideration. Hence it is all fine if T-stack probes are present on any GeneChip®. The scope of this research is to minimize the risk factor of GeneChips® that are being used in diagnostics and other purposes and communicate the effects of thymine-thymine with the scientists of biological world