I am a newcome in bioinformatics and I am studying Dystrophin and how some mutations affect it's structure and may result in Duchenne/Becker's muscular distrophy.
The main problem with that is, since human dystrophin seems to be way too big (3685 aa), there is no solved structure into PDB's database. Some domains are available but all these domains barely add up to 700 aa.
With that, I decided to look into alphafold's database but the full dystrophin is also not available as a complete predicted structure. Actually, they have it all segmented into 1400aa sized fragments with the following logic: frag1 (1-1400), frag2 (200-1600), frag3 (400-1800) and so on.
The main goal now for me would be to combine these overlaing fragments (all into one pdb) and to refine the resulting structure. So what is the best way to do that? Is there something crucial I am not considering?
Massive, multiple domains/subunits, repetitive spectrin-like domains and unstructured or partially disordered regions, especially in its central rod domain ?
João Pessoa
Since the FASTA sequence for human Dystrophin is available, you can predict its full 3D crystal structure using AlphaFold through two main approaches. The first involves de novo prediction without referencing any existing structural models. The second approach allows you to incorporate known fragments from the PDB and AlphaFold database to guide the prediction.
- Badges
- Science topic
- Similar topics
- Medicine
- Disease
- Infectious Diseases