Structural Biology & Proteomics

MD integrates Newton's equations of motion for every atom. Understanding the force field, timestep, and ensemble choices explains why two simulations of the same protein can give different answers.

Docking scores are not binding affinities. Understanding the scoring function (what it approximates and where it breaks down) prevents a very common category of over-interpretation.

Resolution, R-factor, B-factors, clashscore: a structure with poor quality metrics should not anchor your analysis. Learn to read the wwPDB validation report before trusting any PDB entry.

The standard tool for publication-quality structure figures and basic analyses. Free open-source version available. Learn the selection algebra, it's the foundation of all PyMOL scripting.

The most widely used MD engine for protein simulations. Justin Lemkul's lysozyme-in-water tutorial is the canonical starting point, work through it before running your own system.

Fast, free, and widely used for structure-based drug discovery. Understand the receptor preparation steps (hydrogens, charges) before docking: garbage-in applies here hard.

Parse PDB files, extract residues, calculate distances, iterate over chains. The programmatic interface to structure data that you'll need for any batch analysis.

Practical molecular dynamics: the setup, the pitfalls, and what you can extract from trajectories. Relevant directly to GROMACS/AMBER workflows.

How to model assemblies that are too large or too flexible for X-ray crystallography alone. The integrative approach combines cryo-EM, crosslinking, and computational modelling.

Extending structural methods to RNA and RNP complexes: the specific challenges (flexibility, ion interactions) that make RNA harder to model than proteins.

Cryo-EM has become the dominant method for large complexes and flexible systems. Understanding the reconstruction pipeline (particle picking, 2D classes, 3D refinement) is now a core literacy for structural biologists.

No single method gives you the full picture. How to combine cryo-EM, crosslinking-MS, SAXS, and computational modelling into a coherent structural model, and how to report the uncertainty.

The framework for integrative structural biology. Combines heterogeneous experimental data into a structural ensemble with uncertainty quantification.

AF2-Multimer predicts protein complex structures. Interpret pTM and ipTM scores carefully: they indicate model confidence, not binding affinity.

Incorporates experimental data (NMR shifts, mutagenesis, crosslinks) as restraints in docking. More biologically grounded than blind docking when you have any prior knowledge.

The IMP framework in practice: combining cryo-EM density, crosslinking, and other data into a structural model with quantified uncertainty.

How structural differences between species can be exploited for selective drug design. A sophisticated example of how structural knowledge translates into therapeutic strategy.

The structural and computational challenges specific to protein-RNA complexes: flexibility, electrostatics, and the limits of current force fields.

Read it for the architecture and what the pLDDT score actually measures, not just the headline result. The sections on model confidence are particularly important for practical use.

The definitive example of integrative structural biology applied to a large assembly. The methods section is a template for how to combine data types in a principled way.

How structural constraints shape sequence evolution, and what that means for modelling the relationship between structure, function, and selection.