9. Protein structure prediction

Question 1

The structure

Answer 1

is determined by the aa seq.

Question 2

The folding will

Answer 2

correspond to the energy minima

Question 3

Why protein modelling?

Answer 3

• Structure is important for function

- Gap between known sequences and structures is huge

Question 4

Common structures

Answer 4

a-helix
b-sheet
b-turn
random coil

Question 5

Program available for secondary structure prediction

Answer 5

DSSP

STRIDE

Question 6

Different approaches

Answer 6

• Statistical methods
• Knowledge-based methods
• Machine learning
• Consensus method

Question 7

Evaluation 2nd structure

Answer 7

Q3

Sov

Question 8

Q3

Answer 8

fraction correctly predicted residues - Accuracy

Question 9

Sov

Answer 9

Fractional overlap of segments - ability to pick up correct structure

Question 10

How is Q3 and Sov used

Answer 10

They can only evaluate the method itself not your prediction as it looks at already known structures
- one should use both Q3 and Sov for good evaluation

Question 11

Q3 equation

Answer 11

Correctly predicted residues/total residues =Q3%

Question 12

Example of machine learning methods

Answer 12

PSIPRED
PHD
Jnet

Question 13

Membrane topology

Answer 13

look if protein is bound to a membrane or not

Question 14

Common characteristics of TM region (3)

Answer 14

• W or Y at the edges of the membrane
• ca 20 hydrophobic residues inside the membrane
• Positive inside

Question 15

Database to identify TM regions

Answer 15

TOPCON

Question 16

TOPCON

Answer 16

accounts different methods and make a consensus

- gives exact positions of TM region in the bottom

Question 17

3D structure prediction

Answer 17

• More complex than secondary

- Two main approaches

Question 18

Two main approaches of 3D modelling

Answer 18

• Homology modelling

- Ab initio modelling

Question 19

Homology modelling

Answer 19

• use known structure as template
• higher seq similarity -> better prediction
• alignment and template selection is very important

Question 20

Methodology (entire 3D)homology modelling (5steps)

Answer 20

• Identify related structure
• Align target seq to template structure
• Generate “known” backbone and side chains
• Generate loops
• Refine

Question 21

Template selection

Answer 21

• select optimal crystal structure

Question 22

Common approaches for template selection

Answer 22

sequence similarity
homology (both of these: BLAST, Prosite, Pfam)

Fold recognition

Question 23

Fold recognition (threading)

Answer 23

• structure more conserved than sequence
• compare to a known library of folds (CATH, SCOP
• align sequence to a fold
• ENERGY CALCULATIONS
• does not require a similar sequence

Question 24

Loops

Answer 24

• exposed regions are more variable than the protein core
• often important for protein function
• loops longer than 5 residues is hard to model

Question 25

Approaches short, med, long

Answer 25

short - analytical approach
medium - database approach
long - fragment based approach

Question 26

Optimisation of model

Answer 26

• use ENERGY MINIMISATION to fix bad parts
• side chain clashes
• bad peptide bond angles

Question 27

Common errors (of 3D template) 5st

Answer 27

• side chain packing
• distortions/shifts
• bad loops
• misalignment
• bad template

Question 28

Ab initio modelling

Answer 28

• predict structure without any prior knowledge but the sequence
• used when template-modelling fails
• works best for small proteins
• great computational cost

Question 29

CASP

Answer 29

• evaluation of 3D structure methods

- similarity of model to the native structure

Question 30

Function prediction

Answer 30

predict the function
- use results from eg
 * signalP/TargetP
 * 2nd structure pred
 * topology predictions
 * 3D predictions
- use machine learning methods to connect
results usually given as gene ontology terms