Biostatistics research project in Dr.Mar's lab at the Albert Einstein College of Medicine. Benjamin Church and Henry Williams - Work in preparation for publication.
In this paper we explore the potential of a previously underutilized statistical measure - the third moment of distributions - to expand the scope of biostatistics by helping to better understand outliers. In the exploration of this measure we identified methods for ensuring the robustness of analysis, splitting between the tail and non-tail regions of a distribution, and accurately assessing correlation between skewness and expression/methylation. Through our analysis of microarray/RNA-seq expression and DNA methylation data we uncovered multiple significant biological features related to skew as well as significantly more correlation in skewed methylation in the promoter than the body/untranslated regions of genes. Finally we assess the place of this measure in the field going forward and identify areas that could be investigated with the measure.
Applied the analysis of tails of the distributions of expression and methylation datasets from the GDC.