Figure - PMC

Skip to main content

An official website of the United States government

Here's how you know

Here's how you know

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

View full-text article in PMC

. 2018 Jun 21;34(23):3983–3989. doi: 10.1093/bioinformatics/bty476

Search in PMC
Search in PubMed
View in NLM Catalog
Add to search

© The Author(s) 2018. Published by Oxford University Press.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

PMC Copyright notice

Fig. 3. — Parameter selection for FN. The main parameter for FN is the number of principal components of control variation with which to normalize probe quantiles. Screeplots (a, b, d, e) show the metric used to meffil for choosing the optimal number of principal components in ARIES (a, b, c) and GOYA (d, e, f), the amount of probe quantile variation unexplained by the principal components under 10-fold cross validation. The explained variation is mainly due to technical variance as the control probes should not be correlated with biological signal (Supplementary Material). Screeplots (a, d) show the variation without regressing out random effects whereas plots (b) and (e) show the variation after regressing out slide (b) or plate (e) as random effect. Plots (c) and (f) compares true and false positive rates in a downstream EWAS of pre-natal smoking in ARIES (c) and GOYA (f) after normalizing with different numbers of principal components and regressing out slide or plate as a random effect. TPRs were estimated by comparison to associations from a large meta-analysis (Joubert et al., 2016)