Task. Include gene model into h5ad

[michael-kotliar]

When exporting to h5ad file add the following metadata columns: chr, start, end, gene_id. This will allow us to subset anndata object by Gene ID, or select all junctions that overlap the specific interval.

Use it as an example of how to subset by genomic coordinates: scverse/anndata#624 (comment)

Changes should be mainly made in this function:

altanalyze3/altanalyze3/utilities/io.py

Line 134 in b8c06f1

def export_counts_to_anndata(counts_df, location, counts_columns=None, metadata_columns=None, sparse_dtype=None, fill_value=None, strand_coords=None):

Additionally, when annotating junctions, save gene model to the same h5ad file. This gene model should be also updated with the new identified junctions. Need to define in which slot we want to save gene model as it's not something that is structured as var or obs.

[frankligy]

PSI calculation: https://github.com/frankligy/SNAF/blob/main/images/psi_calculation/psi.py
gene model creation: https://github.com/frankligy/SNAF/blob/main/images/gene_model/gene_model.py