skip navigation
Environmental Factor, November 2015

Whole Issue PDF
This issue's PDF is still being created and should be available 3-5 business days after the first of the month. Please check back in a few days.

Epigenetic studies improved by new software from NIEHS

By Kelly Lenox

Headshot of Jack Taylor

“The motivation for developing this tool set was to facilitate analysis of our own methylation data,” said Taylor. “We made the tools publicly available in the hope that other investigators will find them useful and gain better insight into associations between epigenetic changes, disease, and exposure.” (Photo courtesy of Steve McCaw)

Headshot of Zongli Xu

Xu said that ENmix provides a set of flexible and transparent tools for preprocessing EWAS data in a package that is both computationally efficient and user-friendly. (Photo courtesy of Zongli Xu)

An NIEHS team has developed a new tool to improve epigenetic studies looking at DNA methylation, which is a factor that affects how DNA functions. In a paper published Sept. 17 in the journal Nucleic Acids Research, the scientists introduced a novel method for correcting background noise in data from DNA methylation studies.

The researchers showed that the new tool, named ENmix (see sidebar), outperformed other commonly used methods, and they have made the tool available in a user-friendly and freely available software package. By improving epigenetic research, ENmix may help researchers make new insights into the origins of common diseases.

Important difference or background noise?

DNA methylation is a kind of epigenetic change, which means that rather than altering the underlying DNA, it instead affects when genes turn on and off. Epigenome-wide association studies (EWAS) play a key role in the search for links between epigenetic changes and disease. “Complex diseases can be associated with very small differences in DNA methylation profiles,” the authors wrote. Such small differences can be difficult to distinguish from the variability, or background noise, that may be caused by the measurement system used.

The authors said that the measurements of a tool commonly used in EWAS, the Illumina HumanMethylation450 BeadChip, can be affected by many experimental factors. Measurements that result from technical variations rather than useful information, and which generally cannot be reproduced, are known as background noise. Preprocessing the data to remove background noise is an important early step.

“We developed the ENmix background correction method in response to our own research need,” said Jack Taylor, M.D., Ph.D., senior researcher on the team and head of the NIEHS Molecular and Genetic Epidemiology Group. “We have looked at epigenetic change in relation to a variety of diseases, including breast cancer and cleft lip; exposures, such as DES [diethylstilbestrol], genistein, and smoking; and life course, or age. Most involve only small alterations in methylation, so it’s important that the methylation signal be measured accurately and precisely.”

Limitations spark innovation

Zongli Xu, Ph.D., the paper’s lead author and a staff scientist in Taylor’s group, said that existing background correction methods for methylation data had a number of limitations and did not work particularly well. “Either they simply subtracted a fixed amount, or they were based on assumptions about data distributions that do not match those observed in methylation data,” he said.

To develop the software package, Taylor and Xu collaborated with Leping Li, Ph.D., a lead researcher in the NIEHS Biostatistics and Computational Biology Branch, as well as Liang Niu, Ph.D., a research fellow in Li’s group.

“Zongli and his colleagues weren’t content to make do with the technical shortcomings associated with existing approaches,” said Dale Sandler, Ph.D., head of the NIEHS Epidemiology Branch. “Instead, they developed an innovative solution that will benefit other labs working on DNA methylation.”

Greater precision improves methylation studies

The research team demonstrated that ENmix was more accurate and more reproducible than commonly used background subtraction methods. They also showed that the tool substantially reduced technical variations introduced by two different probe designs on the array.

Furthermore, the researchers showed that using ENmix can improve identification of novel disease-associated methylation sites and result in smaller estimates of margins of error.

“We note in particular that ENmix_est [one component of the software suite] has application for the analysis of publically available methylation data sets where background intensity data are not available,” the authors said.

Citation: Xu Z, Niu L, Li L, Taylor JA. 2015. ENmix: a novel background correction method for Illumina HumanMethylation450 BeadChip. Nucleic Acids Res. doi:10.1093/nar/gkv907. [Online 17 Sept. 2015]

"Battling uterine fibroids ..." - previous story Previous story Next story next story - "Jothi earns tenure from ..."
November 2015 Cover Page

Back to top Back to top