Antigens

• Antigen selection, conservation analysis and surface accessibility
• MHC class I and II presentation pathways and the constraints on epitopes
• HLA polymorphism and population coverage as a hard design requirement

Epitopes

• IEDB tools and NetMHCpan-class predictors, with their training-data bias
• Immunogenicity versus binding affinity — predicted binders often do nothing
• B-cell and conformational epitope prediction and its notably poorer accuracy

Chemistry

• RDKit for descriptors, fingerprints and similarity searching
• Similarity-based scaffold hopping and the activity cliffs that defeat it
• Toxicity and off-target screening before a compound is taken further

Integration

• Linking epitope output to chemical screening in one scripted pipeline
• Autoimmunity screening against the human proteome as a safety gate
• A small end-to-end project written up so a reviewer can rerun it