The Bisulphite feature methylation pipeline

The bisulphite feature methylation pipeline is a way to quantitate methylation over a whole feature but using only individual bases whose methylation state is well observed. It then calculates a pecentage methylation for each base and averages these to give an overall value for the feature.

The pipeline assumes that you have methlyation data which uses the standard convention of assigning methylated reads to the forward strand and unmethylated reads to the reverse strand. The reads should generally only be 1bp in length. This is the standard data you would get from importing the output of the bismark methylation extractor via the text import option.

Unlike most other quantitation methods it's possible that this pipeline won't assign a value to some probes in some datasets. In these cases the probe will be given the Not a Number (NaN) value, and will generally be ignored in subsequent calulations and plots.

Options

The options you can set for this pipeline are:

1. Which type of feature you want to quantitate (you can also opt to use the current probe set if you've already made probes)
2. The minimum level of observation required to include a base in the methylation calculation. This is simply the depth of count over that base. This value should be a fair reflection of the overall depth of coverage in your data. If you only have a 3X average coverage of all bases then there's no point setting this value to 5 since you'll just throw away all of your data.
3. The minimum level of observation for a feature - ie how many individual bases need to have been called to make a call for the feature as a whole.
4. How to combine the measures for a given feature. You can choose to report the mean, median, max or min value from the set of individual base measures.