Sometimes bioinformatics can seem a bit mystical, particularly when a tool arrives that makes you wonder what you were wasting your time doing before it. ChIP-seq is no stranger to this world and as a driver of a large proportion of epigenomic experimentation we’re also no stranger to the challenges of processing ChIP-seq data and peak calling.
One of the biggest challenges currently faced by ChIP-seq is the calling of differential peaks between two cellular conditions. Most bioinformatic software can only analyze one ChIP-seq signal at a time, and thus can’t let you know when DNA protein interactions differ in your conditions of interest. In attempt to overcome this difficulty a few programs have come about that combine peak callers and statistics for detecting differential peaks, but most fail to catch the finer details of protein-DNA interactions. However, now there are two new players to help you overcome differential peak calling obstacles:
ODIN: One ChIP-Seq Program to Call Them All
Researchers from RWTH Aachen University (Germany) touch on their Norse roots, with a ‘father of the gods’ that seems to be able to do it all. It can call your peaks as well as let you know when there are differential peaks between your conditions, all in a integrated framework that gives you everyone’s favourite metric; the p-value. The team doesn’t just stop there though, they also introduce an evaluation methodology to compare ODIN to its competitors. Their empirical tests that utilize several ChIP-seq experiments show that ODIN almost always trumps the competition.
Give ODIN a run with your data and check out the full story in Bioinformatics, November 2014
ChIP-Rx: The Differential Peak Detection Powers of Reference Normalization
Researchers from Syros Pharmaceuticals in Massachusetts have also come up with some ChIP-seq sorcery. But in their case, the motivation is the fact that there isn’t an empirical normalization strategy to properly compare and call peaks across different conditions and samples in ChIP-seq experiments. Their solution? ChIP-Rx, it lets you compare your peaks qualitatively by using reference chromatin and is adept at discovering graded changes in response to exposures that would be not be called in other normalization strategies.
The paper also offers a solid demonstration of ChIP-Rx in action and brings home the point that reference normalization of ChIP-Seq experiments will help drive forward the discovery of disease related histone mod patterns.
Check out all the details on the power of reference chromatin and on how traditional normalizations can obscure epigenomic differences in Cell Reports, November 2014