Predict MANE related structures

[St3451]

Find a way to predict MANE-related structures.

This would solve the issue of missing the structures of some MANE transcripts. Once we have all the structures, we could use it straightforwardly in a pipeline like IntOGen.

1. Find a way to get protein sequence: we do have an associated Uniprot ID (either provided by the AlphaFold DB MANE release or by the MANE dataset itself.. however, these entries are often missing in Uniprot), we could get it from there.

2. Find a way to run AlphaFold2: we could use the nf-core pipeline and run it either on the cluster or on the cloud (bbglab org from seqera platform?)