What is the difference between secondary structure assignment and prediction?
• Assignment: based on 3d structure, eg pdb coordinates, assign structure elements
• Prediction: based on sequence, predict secondary structure elements
What distinguishes loops from alfa helices and beta sheets?
• Very disordered
• High flexibility
What are the main differences between α-helices and β-sheets?
• Alfa helices:
• local backbone interactions
• long sequence regions
• Beta sheets:
• backbone interactions can be local or distant
• short sequence regions
Amino acids have preferences for secondary structure types. Give some examples/key rules.
• Bulky residues with a Cβ branch tend to favor β-sheets.
• Smaller amino acids tend to favor α-helices and loops
• G and P tend to occur at the ends of a helix and in loops
• (glycine and proline)
How do you calculate a propensity?
• Calculate the fraction of a specific residue in a secondary structure type
• Calculate the fraction of all residues in a secondary structure type
• Divide the two to get the propensity
• Propensity < 1: secondary structure type not favored by residue
• Propensity > 1: secondary structure type favored by residue
What is the difference between classification and regression?
See slide.

Secondary structure  can be viewed as a classification problem
Explain the general concept of viewing secondary structure assignment as a classification problem.
See lslide
What is the concept of double loop cross validation?
See slide
Explain the difference between parameters and hyperparameters.
• Parameters: parameters of the model (e.g. variable weights)
• Hyperparameters: parameters not directly in the model (e.g. number of cross validation loops, size of training and test set)
Explain "neighbourhood matters!" in the context of secondary strucrure assignment.
• Residues are not independent of one another
• Hydrophobicity patterns depending on location of secondary structure element
• If the residues at positions 5 and 7 of a sequence are both in an alpha-helix, the residue at position 6 is also likely to be an alpha-helix
Chou-Fasman is an example of a window based approach. Explain.
• Score every amino acid based on α-helix and β-sheet propensity
• Look for series of ≥4 amino acids which for example have high scores for α-helix
• If average α-score > β-score →accept as α-helix

Slinding windows are also used eg GOR method
• A unique study and practice tool
• Never study anything twice again
• Get the grades you hope for
• 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.