331

solution by locally changing the current solution to find a better solution from the neigh­

borhood under consideration.

Simulated annealing strategy: In each case, to solve the traveling salesman problem as

well as possible.

Question 8.3

• Monte Carlo

• simulated annealing (since pointing to the protein folders, e.g. at SWISS-MODEL

there is a refinement, which should be in the direction)

• evolutionary strategies

• genetic algorithm (is implemented in YANASquare, mention that then)

• optimizer (steepest descent, also mention the procedure for YANAvergence, the

Broyden-­Fletcher ...).

Question 8.4

A difficult computational problem is a bioinformatics problem in which many possibilities

lead combinatorially to an exponential growth of possibilities, e.g., the traveling sales­

man’s problem of traveling to numerous cities along the most optimal route possible.

These exponentially complex problems with very, very long computation time for system­

atic trial and error (longer than the universe exists, etc.) are contrasted with easier prob­

lems where the computation time grows only polynomially (P-problems), e.g. quadratically

or cubically with the length of the query, such as the sequence length. However, almost all

interesting bioinformatics problems are combinatorial (e.g. protein folding or possible

protein complexes). It has also been shown that they are all analogous to the traveling

salesman problem, i.e., they require non-polynomial computation time, are NP-hard.

20.9

Complex Systems Behave Fundamentally in a Similar Way

Question 9.1

The behavior of ordered system is predictable and exactly describable for the whole

period. Random systems are unpredictable in the short term, but the outcome space can be

predicted (such as a dice, can only be one to six). In addition, there are chaotic systems that

can only be described exactly over short periods of time, but remain within fixed limits

(attractor) over the long term.

Question 9.2

Here we have learned about numerous systeming ingredients in the book: Modular units

(nucleic and amino acids) have interactions and in turn form complexes and networks (e.g.

feedback or feedforward loops), from which filaments, organelles, tissues and ultimately

20.9  Complex Systems Behave Fundamentally in a Similar Way