Protein Folding Problem

Question 1

What is the protein folding problem?

Answer 1

The protein folding problem is the question of how the amino acid sequence of a protein dictates its structure.

Question 2

In what year did Science name the protein folding problem as one of the biggest unsolved problems in science?

Answer 2

In 2005, Science named the protein folding problem as one of the 125 biggest unsolved problems in science.

Question 3

What are the three problems of protein folding?

Answer 3

• The folding code
• The computational problem
• The kinetic question

Question 4

What is the folding code in protein folding?

Answer 4

The thermodynamic question of how a native structure results from the interatomic forces acting on an amino acid sequence.

Question 5

What was the predominant view of the protein folding code before the mid-1980s?

Answer 5

It is the sum of many different small interactions expressed through secondary structures and local in the sequence.

Question 6

Newer view to the folding code

Answer 6

• There’s a dominant component to the folding code (hydrophobic interaction)
• It is distributed both locally and non-locally in the sequence
• The native secondary structures are more a consequence than a cause of folding forces

Question 7

How did a different view of the folding code emerge?

Answer 7

Through statistical mechanical modelling

Question 8

What is the computational problem in protein folding?

Answer 8

How to predict the native structure of a protein from its amino acid sequence using computers.

Question 9

What was the first major milestone in solving the computational problem?

Answer 9

The supercomputer simulation by Duan and Kollman in 1998
- The 36-residue villian headpiece from an unfolded to a folded state

Question 10

What is Folding@home, and what is its goal?

Answer 10

A distributed computing project
Designed to perform computationally intensive simulations of protein folding, misfolding, and related diseases.

Question 11

How does Folding@home work, and how was it recognized by the Guinness Book of Records?

Answer 11

• Works by people all over the world downloading and running software on their own PC/Games console, creating one of the world’s largest supercomputers.
• In 2007, the Guinness Book of Records recognized Folding@home as the most powerful distributed computing cluster in the world.

Question 12

What is Foldit, and how is it different from Folding@home?

Answer 12

• Takes the concept and adds a human problem-solving element.
• instead of donating idle CPU cycles to perform scientific research it presents unfolded proteins to the player in form of puzzles

Question 13

What is the most interesting feature of Foldit, according to David Baker?

Answer 13

It incorporates competition into the game between gamers and actual research groups, allowing for a faster and more collaborative approach to solving problems.

Question 14

What is CASP and when was it started?

Answer 14

• Critical Assessment of Techniques for protein Structure Prediction
• bi-annual event started by John Moult in 1994

Question 15

What is the purpose of CASP?

Answer 15

To predict unknown protein structures, given only the amino acid sequence

Question 16

How are the computer-based structure predictions in CASP evaluated?

Answer 16

By comparing them to experimentally resolved structures and assigning an accuracy score;
the Global Distance Test - Total Score (GDT_TS or GDT) is used to measure the similarity between the predicted and resolved structures.

Question 17

What is AlphaFold2?

Answer 17

• AlphaFold2 is an artificial intelligence system developed by DeepMind
• Won CASP14 in 2020.

Question 18

What was AlphaFold2’s performance in CASP?

Answer 18

• GDT >90 for ~2/3 of the target proteins
• GDT >80 for almost 90% of the targets
  results considered competitive with experimental approaches.

Question 19

When was DeepMind founded and who acquired it?

Answer 19

• Founded in 2010
• Acquired by Google in 2014
• Owned subsidiary of Alphabet Inc. in 2015

Question 20

How does DeepMind’s approach to AI differ from other well-known AIs like IBM’s Deep Blue or Watson?

Answer 20

• They are not pre-programmed
• They learn from experience using only raw pixels as data input, based on a system.

Question 21

What is DeepMind’s mission?

Answer 21

“solve intelligence to advance science and benefit humanity.”

Question 22

What is the kinetic question?

Answer 22

How can proteins fold so fast

Question 23

What is Levinthal’s paradox?

Answer 23

The puzzle of how proteins can fold so quickly
despite the very large number of possible conformations they can adopt,
which would take longer than the age of the universe to sequentially sample.

Question 24

When was Levinthal’s paradox first noted?

Answer 24

In 1968 by Cyrus Levinthal

Question 25

What is the folding funnel hypothesis?

Answer 25

A version of the energy landscape theory of protein folding which assumes:
that a protein’s native state corresponds to its free energy minimum under the solution conditions encountered in cells.

Question 26

What does the folding funnel hypothesis suggest

Answer 26

That the folding process is guided by a funnel-shaped energy landscape
where the native state is located at the bottom of a deep free energy minimum with steep walls.

Question 27

Are energy landscapes typically smooth or rough, and why?

Answer 27

Rough
with many non-native local minima in which partially folded proteins can become trapped.

Question 28

What is entropy?

Answer 28

• The measure of disorder or randomness in a system.
• A measure of a system’s internal thermal energy that is unavailable for doing work.
• As a system takes on energy, its entropy increases, and it tends to move towards a high-energy, disordered, and unstable state.

Question 29

Why are chaperone molecules important for protein folding?

Answer 29

They
* isolate individual proteins to prevent interruptions from interactions with other proteins or
* help to unfold misfolded proteins, giving them a second chance to refold properly.

Without chaperones, proteins may aggregate and form insoluble amorphous aggregates.

Question 30

What is the risk of protein aggregation?

Answer 30

Aggregated proteins are associated with various diseases.
Such as
* Creutzfeld-Jakob disease - prion
* Alzheimer’s disease - amyloid
* Parkinson’s disease. -intracytoplasmic

Question 31

How does a protein chain avoid traps and hills during folding?

Answer 31

One microscopic mechanism:
Zipping and assembly (ZA)

Question 32

What is the ZA mechanism?

Answer 32

• Proteins can fold quickly because they don’t reach all their degrees of freedom at the same time.
• Proteins fold over a wide range of timescales.
• On the fastest timescales (picoseconds to nanoseconds), different small peptide pieces of the chain explore local conformations independently of other such pieces.
• Local structure forms and then grows (zips) to include increasingly more surrounding chain.
• Multiple pieces may then assemble together on slower timescales.

Question 33

What are the steps in the ZA folding route for Protein G?

Answer 33

1. The chain is parsed into many short, overlapping fragments.
2. The fragments are sampled by replica exchange molecular dynamics to identify stable hydrophobic contacts.

Question 34

What are the two ways in which fragments are further processed in the ZA mechanism for Protein G?

Answer 34

1. Grown or zipped through iterations of adding new residues, sampling, and contact detection,
2. Assembled together pairwise using rigid body alignment followed by further sampling until a completed structure is reached.

Question 35

What answer does the ZA mechanism provide?

Answer 35

• Plausible answer to Levinthal’s kinetic protein folding problem
• Shows why proteins dont need supercomputers to guide them to their native structures.