Modeling X Chromosome Data Using Random Forests: Conquering Sex Bias

Intersections of machine learning and epidemiological methods for health services research

The field of health services research is broad and seeks to answer questions about the health care system. It is inherently interdisciplinary, and epidemiologists have made crucial contributions. Parametric regression techniques remain standard practice in health services research with machine learning techniques currently having low penetrance in comparison. However, studies in several prominent areas, including health care spending, outcomes and quality, have begun deploying machine learning tools for these applications. Nevertheless, major advances in epidemiological methods are also as yet underleveraged in health services research. This article summarizes the current state of machine learning in key areas of health services research, and discusses important future directions at the intersection of machine learning and epidemiological methods for health services research.

Statistical learning approaches in the genetic epidemiology of complex diseases

In this paper, we give an overview of methodological issues related to the use of statistical learning approaches when analyzing high-dimensional genetic data. The focus is set on regression models and machine learning algorithms taking genetic variables as input and returning a classification or a prediction for the target variable of interest; for example, the present or future disease status, or the future course of a disease. After briefly explaining the basic motivation and principle of these methods, we review different procedures that can be used to evaluate the accuracy of the obtained models and discuss common flaws that may lead to over-optimistic conclusions with respect to their prediction performance and usefulness.

Efficient Statistical Method for Association Analysis of X-Linked Variants

BACKGROUND/AIMS

Unlike the gene-poor Y chromosome, the X chromosome contains over 1,000 genes that are essential for viability of cells. Females have 2 X chromosomes, and thus female X-linked gene expression would be expected to be twice that of males. To adjust this imbalance, one of the 2 X-linked genes is often inactivated, and this is known as X-chromosome inactivation (XCI). However, recent studies described that a gene can be nonrandomly selected for inactivation from 2 X-linked genes and that XCI is not observed in some X-linked genes. Since this complex biological process has prevented efficient statistical association analyses, we propose a new statistical method against this uncertain biological process.

METHODS

The proposed method consists of 2 steps. First, p values for various biological processes are calculated and then combined into a single p value with the modified Fisher method and a minimum p value.

RESULTS

Our simulation results show that the proposed method is generally the most statistically efficient and is not sensitive to the unknown biological model.

CONCLUSION

Therefore, we can conclude that the proposed approaches are robust against the various XCI processes for testing the association of X-linked single nucleotide polymorphisms with the disease of interest and the proposed method is a practical solution.