RNA types

[eboileau]

Aims/objectives.

The nomenclature for RNA types (table rna_type) is fixed e.g.

+---------+-------------------+
| id      | name              |
+---------+-------------------+
| circRNA | circular RNA      |
| mRNA    | messenger RNA     |
| miRNA   | micro RNA         |
| rRNA    | ribosomal RNA     |
| snRNA   | small nuclear RNA |
| tRNA    | transfer RNA      |
+---------+-------------------+

This means that anyone can eventually upload modification data associated with any of these RNA types. According to the specification,

a given dataset or bedRMod file can contain more than one modification, as reported in column 4 (MODOMICS short name), but this should be for the same RNA type.

otherwise it becomes practically impossible to search records, at least the way the model is designed.

A clear and concise description of todo items.

• Do we actually need a special treatment for the RNA types other than tRNA? i.e. all of these can be annotated with Ensembl exactly as for mRNA, as far as I known. The only problem is that in principle this would require different bedRMod files (dataset) for different RNA types, whereas generally mapping includes all of mRNA, miRNA, rRNA (incl. MT), etc. In practice it may be difficult to differentiate mRNA from the other common small RNAs...
• tRNA should definitely be handled separately, and in general anyway, mapping of mRNA/other RNAs will not include tRNA, unless I'm mistaken. On this, see https://github.com/dieterich-lab/scimodom/discussions/69, but I have no clue how this can be implemented in practice. And do tRNA bedRMod files are the same (genomic coordinates, etc.)?

[eboileau]

The terminology is now fixed to

+------+---------------------+
| id   | name                |
+------+---------------------+
| tRNA | transfer RNA        |
| WTS  | whole transcriptome |
+------+---------------------+

It is built around the concept of sequencing method rather than the biological definition of RNA species. WTS thus stands for mRNA, ncRNAs, etc.

I started to work on integrating tRNAs...

We can eventually add rRNA, but unless I'm mistaken, this would not require changes, as they will be annotated with Ensembl and be handled as WTS.