Re-tried MACAU kicking out the one poor quality sample. Got a few more meaningful loci to look at, but all had Beta values that were indistinguishable from 0.
Just like the 6 sample version, only with 5!
files were too big/disparate for R to handle, sqlshare too big of a pain, so I did it the hard way!
I'm sure there's an easier way to do this, but it seemed to work.
Just needed to change the flagW and flagC arguments to 0 and 16 respectively.
BSMap comparison for trimmomatic trimmed files
Re-ran Bismark with re-trimmed files. No measurable change in mapping rates.
Re-ran trimming step using Trimmomatic and TruSeq adapter sequences specified.
Bismark run on C. virginica oil exposure samples. Super low mapping rate (sub 10%) which is strange...
Fastqc, trimming, and re-fastqcing
Moving and checking files, prepping the genome.
This time it works!
Testing upload for R Notebook
or: Why scientists shouldn't be allowed to name things.
Tried an ANOVA on mean methylation for Day 10 samples by treatment. Nothing really interesting came of it. On to day 135!
Forgot to upload this. Super time/space intensive but its nearly done after a lot of hard drive shuffling!
Making a file of percent methylation to start looking at the geoduck methylome.
Looked at the combined Day 10 and Day 135 samples, with the inclusion of a time covariate.
These are the day 145 runs, with the corrected methylation cutoff logic. Macau was run outside of R-studio, to facilitate running multiple instances concurrently.
So I noticed that I was inadvertently trimming all methylation counts less than 10, as well as the total counts. I should not have done that. The good news is this nets us 267 DMRs as opposed to 41! The bad news, for some reason it makes MACAU run glacially slow.
Pseudocode description on how to combine multiple SNP files in to a single merged file, and then generate a relatedness matrix.
Running Methyl Extract on a single sample .sam file at a time.
Rough protocol outline after meeting with Micah at DNR in Olympia.
CpG O/E plot for Geoduck transcriptome, try 2.
Forest Plot code for Geoduck transcriptome methylation
Testing R notebook functionality with some Methylkit output from Steven's Oly samples.