Hi everyone,

I’m currently working on designing a multi-epitope vaccine and have used AlphaFold to predict the 3D structure of the protein sequence I designed. However, I’m encountering an issue: the confidence score provided by AlphaFold is relatively low for my sequence, especially considering that I designed it myself and didn’t download it from a database like UniProt or PDB.

Is it normal for AlphaFold to provide a low confidence score when predicting structures for custom-designed sequences (e.g., multi-epitope vaccine constructs)? Are there any recommendations on how to improve the confidence score, or any additional steps I should take to refine the structure prediction?

Thanks in advance for your help!