Meta-Analytic Operation of Threshold-independent Filtering (MOTiF) Reveals Sub-threshold Genomic Robustness in Trisomy

Roy Siegelmann,Hava Siegelmann
DOI: https://doi.org/10.48550/arXiv.2302.00772
2023-02-02
Abstract:Trisomy, a form of aneuploidy wherein the cell possesses an additional copy of a specific chromosome, exhibits a high correlation with cancer. Studies from across different hosts, cell-lines, and labs into the cellular effects induced by aneuploidy have conflicted, ranging from small, chaotic global changes to large instances of either overexpression or underexpression throughout the trisomic chromosome. We ascertained that conflicting findings may be correct but miss the overarching ground truth due to careless use of thresholds. To correct this deficiency, we introduce the Meta-analytic Operation of Threshold-independent Filtering (MOTiF) method, which begins by providing a panoramic view of all thresholds, transforms the data to eliminate the effects accounted for by known mechanisms, and then reconstructs an explanation of the mechanisms that underly the difference between the baseline and the uncharacterized effects observed. As a proof of concept, we applied MOTiF to human colonic epithelial cells, discovering a uniform decrease in gene expression levels throughout the genome, which while significant, is beneath most common thresholds. Using Hi-C data we identified the structural correlate, wherein the physical genomic architecture condenses, compactifying in a uniform, genome-wide manner, which we hypothesize is a robustness mechanism counteracting the addition of a chromosome. We were able to decompose the gene expression alterations into three overlapping mechanisms: the raw chromosome content, the genomic compartmentalization, and the global structural condensation. While further studies must be conducted to corroborate the hypothesized robustness mechanism, MOTiF presents a useful meta-analytic tool in the realm of gene expression and beyond.
Quantitative Methods
What problem does this paper attempt to address?