Bioinformatics 17: Structural 2

Question 1

What is a Ramachandran Plot?

Answer 1

Ramachandran Plot is a way to visualize dihedral angles ψ against φ of amino acid residues in protein structure. Ramachandran recognized that many combinations of angles in a polypeptide chain are forbidden because of steric collisions between atoms

• Method of predicting protein structure

Question 2

What amino acid sequence patterns suggest Beta-strand structure?

Answer 2

Short runs of conserved hydrophobic residues suggest a buried Beta-strand

i, i + 2, i + 4 patterns of conserved hydrophobic residues suggest a surface beta-strand

Question 3

What amino acid sequence patterns suggest Alpha-helix structure?

Answer 3

i, i + 3, i + 4 and i + 7 (and variations of this pattern) of conserved residues suggest a-helix with one surface facing solvent

Question 4

Why are alpha-helices much easier to predict than beta-strands? What about trans-membrane helices?

Answer 4

Alpha-helices are generally longer -> more information is available -> higher confidence

Trans-membrane a-helices are even easier to predict, as the membrane width acts as a frame of reference

Question 5

What is ab initio modelling?

Answer 5

Ab initio- or de novo- protein modelling methods seek to build three-dimensional protein models “from scratch”

i.e. via physical properties and interactions, not previously known biological characteristics / solved structures

Question 6

What is in silico folding and what is the problem with using it for larger proteins?

Answer 6

Folding via computer simulation

-> computer must calculate the interactions of a single amino acid with every other amino acid within the protein -> non-linear scaling, requires unrealistic amounts of computing power at anything more than ~100aas

Question 7

What is homology modelling?

Answer 7

‘Comparative modeling’

• Construction of an atomic-resolution model of the “target” protein from its amino acid sequence + experimental 3D structure of a related homologous protein (the “template”).
• Relies on the identification of one or more known protein structures likely to resemble the structure of the query sequence, and on the production of an alignment

Question 8

Steps of homology modelling?

Answer 8

Identify known 3D structure (template) similar to unknown structure (query)

Produce alignment

Use known structure to produce homology model

Refine model

Validate model

Question 9

How do you identify a template in homology modelling?

Answer 9

Use blast to search PDB database to find templates

High similarity + alignment = good

Question 10

Modelling software?

Answer 10

MODELLER (single step approach)

SWISSMODEL (fragment based, stepwise approach)