Counter, DGE, Plotter: optionally break down counts per-feature using custom tags

[AlexTate]

We would like to allow users to break down counts per-feature by defining a tag on a per-rule basis.

• [x] Counter An additional column will be added to feature_counts.csv which contains the tag value for features which matched a "tagged" rule. Counts that result from "tagged" rules will be tallied separately and shown on the corresponding row. For example, if 100 reads were assigned to a feature via a non-tagged rule and 50 reads were assigned via a tagged rule, then feature_counts.csv would contain:

  Feature ID	...	Tag	sample_rep_N
  example_feat	...		100
  example_feat	...	example_tag	50

  Counts resulting from "non-tagged" rules will be pooled as they are now (and they would appear on rows which lack a Tag column value in feature_counts.csv). If multiple rules define the same tag, then counts from these rules will be pooled under the same tag for all matching features.

• [x] DGE During DGE analysis, tiny-deseq.r will concatenate feature IDs and tags, set the dataframe index to this (optionally) tagged feature ID, drop the Tag column, then perform DGE as it currently does. This will allow for separate DGE analysis for counts originating from "tagged" rules on a per-feature basis. Once DESeq2 is complete, "tagged" feature IDs will be split so that original feature IDs and tags are in separate columns once again.

• [x] Plotter Plots for scatter_by_dge_class and class_charts will also break down counts per class-tag pair

[AlexTate]

Update: first drafts of changes in Counter and DGE are complete.

Plotter will require some tricky processing of the Feature Class column to ensure that class-related charts properly represent tagged class counts for features which have multiple associated classes and match multiple tagged rules.