Update the annotation files for scRNA analysis

[chris-kreitzer]

06/09: match the ortho_table (Juan) with old NVE annotations (Alison)

Fact is:

• A unique NV2_ID matches several NVE_IDs (e.g. NV2.14867 matches NVE10, NVE13262, etc.)
• Several NV2_IDs don't have a corresponding NVE annotation
• NV2_IDs do match NVE_IDs several times (based on different JGIProtID, ENSEMBL_ID, etc.)

To resolve this issue, I propose the following:

• stick with the old list; and just work with the corresponding NV2 annotations;
• completely revise the annotation and forget about the NVE annotations (annotations, genes, TF_List)
• speak with Juan and Alison about this; right now, I am just working with the old annotations where we have corresponding new?? NV2_IDs

[chris-kreitzer]

06/13: update of annotations: NVE to NV2

• we stick with the 'old annotation table' (Alison)
• we concentrate on vienna_nv2.confident.map (made by Juan, several shorter NVE transcripts match to one NV2 model) - Juan, explained the logic behind it ~ confidence map, where we can get sure, that mappings fit

  Three scenarios

1. if NVE doesn't have corresponding NV2 --> discard this annotation
2. if 1 NVE does have multiple (max 4) NV2's, the NVE annotation (gene_ontology_pfam) gets just duplicated and assigned the new ID (NV2)
3. if 1 NV2 map to several NVE's, take all the gene_ontology_pfam annotations from all the NVE's and assign to unique NV2 (see script)

This approach ends up with 17,811 features!