Protein folding — energy funnel and Anfinsen (3D)
About this app
A real protein chain folds before your eyes: hydrophobic collapse, helices and sheets, energy funnel and Levinthal calculation. Heating unfolds it, cooling folds it again.
With explain mode: an AI tutor walks you through the app, points out what matters in the scene and answers your questions.
Sequence determines shape
The order of amino acids holds the information for a protein’s 3D structure (Anfinsen). Water-repelling residues pack together inside, and hydrogen bonds in the backbone form helices and sheets. If a chain tried every shape at random, even a small 35-residue protein with 3³⁵ ≈ 5·10¹⁶ shapes at 10¹³ tries per second would need over an hour, yet real proteins fold in micro- to milliseconds (Levinthal’s paradox). The answer is the energy funnel: each correct contact lowers the energy, and many paths lead to the same structure.
What the app shows
Three proteins from the Protein Data Bank fold in a simulation: the villin headpiece HP35 (35 residues), Trp-cage (20 residues) and the B1 domain of protein G (56 residues). It runs a Gō model with one bead per amino acid and Langevin dynamics; the model knows the target structure and does not predict it. Read-outs for native contacts Q, deviation from the target and radius of gyration, a funnel with the trace of each run and the Levinthal estimate accompany the folding.
Things to try
Heat the chain above its melting point: the fold falls apart but the chain itself stays intact. Cool it down and use “Unfold again” for more runs, each leaving its own trace in the funnel. Switch between primary, secondary and tertiary level, colour by property, structure or chain, and pick single residues in the sequence panel to see their contacts.
Frequently asked questions
- What is Levinthal’s paradox?
- Cyrus Levinthal pointed out that a protein chain has an enormous number of possible shapes. With just three options per amino acid, a 35-residue chain already has about 5·10¹⁶ shapes, and a random search would take far longer than the measured folding times of micro- to milliseconds. The resolution is that proteins do not search at random; they roll downhill in a funnel-shaped energy landscape.
- What happens when a protein is denatured?
- When heated, hydrogen bonds and other weak contacts break and the 3D structure comes apart. The peptide bonds, and with them the primary structure, remain intact. Small proteins often refold on cooling, whereas large ones frequently clump together for good.
- What is an energy funnel in protein folding?
- The energy funnel is a picture of a protein chain’s energy landscape. The wide rim holds a huge number of disordered, high-energy shapes, and the narrow bottom holds the single native structure. Every correctly formed contact lowers the energy, so many different paths lead down to the same goal.
- What can I do in this app?
- Choose one of three real proteins, set the temperature, start new folding runs and follow Q, deviation and radius of gyration. A funnel traces the path of each run, and the Levinthal panel compares it with a random search. The sequence panel highlights individual amino acids and their contacts.
Subject: Biology & neuroscience
More from Biology & neuroscience
- Aquaporin — water in single file (3D)
- ATP synthase — the cell's motor (3D)
- Eye — eye muscles and vision (3D)
- CRISPR-Cas9 — the gene scissors (3D)
- The retina and blood sugar (3D)
- Fibonacci cactus — one angle, many spirals (3D)
- Gait — walking from real measurements (3D)
- Haemoglobin — cooperativity (3D)
A VisuApp by heyprof: interactive, free in your browser, no sign-up. What is a VisuApp? · All apps