4.5 Structure Prediction Landscape

Structure Prediction Landscape (Module 4: The Bioinformatician's Toolkit). Key concept: AlphaFold & prediction tools. Explain what a structure predictor does, and what its confidence scores are actually scoring.

• Explain what a structure predictor does, and what its confidence scores are actually scoring
• Read pLDDT as per-residue confidence and identify which parts of a model to trust
• Read a PAE matrix to judge whether a domain–domain orientation is reliable
• Compare the current methods — AlphaFold2 as the MSA-driven workhorse, AlphaFold3 and its open successors, ESMFold for speed — and choose for the problem in front of you
• Recognise that low pLDDT often marks genuine disorder rather than a failed prediction
• Interrogate a prediction rather than accepting it