Identification of Hypermethylated Loci in RdDM-deficient Genetic Backgrounds

The RNA-directed DNA methylation (RdDM) pathway in Arabidopsis thaliana is responsible for de novo placement of methylation on cytosines. Particularly in the CHH (H = A, T, or C) sequence context (asymmetric methylation). However, some subset of loci actually become hypermethylated when the RdDM machinery is dysfunctional, particularly when RNA polymerase IV subunits are knocked out. I would like to look for more loci which become hypermethylated in RdDM deficient genetic backgrounds and see if these loci have anything in common with regard to their annotated functions.

Concept Map v 2.0:

Code:

Code will be accessible through my github.

Documentation:

Programs

Science

Updates:

Update - 2-23-15

Update - 3-3-15

Update - 3-10-15

Update - 3-24-15

Update - 3-31-15

Update - 4-7-15

Update - 4-14-15

Update - 4-21-15

Update - 5-1-15

Completed Items:

Completed Items

Red Flags/Issues:

3/9/15 - Tested my HPC access.That's all in working order. No Bismark on UA's HPC though.

3/10/15 - I'm abandoning my idea to look at the AGO4 knockout background. The data exists, but is from a different group. This raises questions about how comparable it is to the other datasets. Also, I already have plenty of data to analyze (some might argue too much). I do have a Pol IV + Pol V KO double mutant though. I will run that through the analysis if there is time instead.

3/23/15 - Writing a python script to parse Bismark output files is probably going to be very challenging, so I'm investigating code that's freely available and whether I can either figure it out and replicate something similar, or us it as is.