SWF: Identification

[ninsch3000]

Notes to reviewers

most careful review needed on file compute_metrics.py, as that is where the identification specific calculations are introduced. I used #239 as inspiration, but ended up implementing most of it myself. Note that the required changes to utils/apaeval/ had been already done in my previous PR #415.

Fixes #175 #300 Makes PR #239 obsolete, closing that one manually

Checklist

• [x] I have performed a self-review of my own code
• [x] I have commented my code, particularly in hard-to-understand areas
• [x] I have updated corresponding READMEs (if applicable)
• [x] My code follows the templates/style guidelines of the repository
• [x] In- and output formats comply with APAeval specifications
• [x] No parameters or file names are hardcoded
• [x] Results, logs or other output is not commited to the repository
• [x] I have tested my code

[ninsch3000]

Optional metrics missing

Several metrics from the identification specification haven't been implemented yet:

• [ ] metric: Percentage of genes with correctly identified number of PAS
• [ ] metric: PAS assigned to different genomic features: terminal exons, exons (excluding terminal exons), introns, intergenic regions (less and more than 1kb downstream of 3'-terminal exon)
• [ ] Plot: 3. Bar plot visualizing percentage of genes with correctly identified number of PAS
• [ ] Optional plot: 1. 2D line plot visualising Sensitivity (Recall) and Precision in the form of Precision-Recall curve.
• [ ] Optional plot: 2. Grouped bar plot visualizing proportions of PAS assigned to different genomic features in one plot.
• [ ] Optional plot: 3. Bar plot visualizing proportion of PAS matched to multiple ground truth sites.

Not all suggested plots are created either.
PLEASE updated this comment as needed. (more checklist items, etc)