The Simulation Argument meets Evolutionary Informatics

From the Simulation Argument:

Unless we are now living in a simulation, our descendants will almost certainly never run an ancestor-simulation.

I think it would be interesting to run a few ancestor-simulations. What is our simulation likely to be like then? What about an evolutionary algorithm?

In evolutionary informatics, two paramaters are needed for evolutionary algorithms. Heritability and Selection.

Heritability implies the following:

1. "Parents" give rise to "offspring".
2. Traits from "parents" are passed on to "offspring".
3. Each "offspring" from a "parent" signifies a new generation.
4. Variation between generations may or may not occur.

Selection implies that certain traits that are not on a fitness landscape will not be selected.

Let's look at Autodock as an example and how it relates to evolutionary informatics. Autodock employs a genetic evolutionary algorithm in order to try and predict the orientation of a ligand within a protein.

The ligand is the heritable structure. (A ligand is any structure that binds to a protein, e.g. a therapeutic molecule)
The protein is the fitness landscape.
The genetic evolutionary algorithm provides the variation and selection parameters.
Consider the following diagram:
Figure 1: A) Basic lay out of memetic algorithms. A population of individuals is randomly seeded with regard to fitness (initialized). The individuals are randomly mutated and their fitness is measured. Individuals with optimal fitness are further mutated until convergence of a local optima is reached. The process is carried out for the entire initialized population. The global optima is selected from the various local optima. B) Fitness landscape with local optima (A, B and D) and a global optima (C). In a memetic algorithm, the initial population of individual are randomly seeded and can be viewed as any of the arrows indicated in the figure.

A few important aspects from the figure:

1. Fitness depends on the phenotype.
2. Fitness (in the case of Autodock) is the capability of the ligand phenotype to bind and stay bound to the protein.
3. The parameters for succesful binding are many. For Autodock, the following are included:

• Van der Waals interactions
• Electrostatic interactions
• Desolvation,
• Hydrogen bond interactions
• Torsional free energy
• Conformational interactions

If certain parameters (above) are not on a fitness landscape for a certain ligand phenotype such as the absence of hydrogen bonds at a particular area of the protein, such a trait will not aid in ligand binding for a particular ligand with hydrogen bonds. Therefore,hydrogen bonding (as a trait) will not be on the fitness landscpe and is thus not a selectable trait.

Autodock uses a Solis & Wets search algorithm to probe the fitness landscape of a particular protein (See figure below).

Autodock Solis & Wets algorithm.

The surface of a protein is where the binding of the ligand will occur, thus 3-dimensionally, the fitness landscape would resemble something like this:

Rapamycin ligand bound to the mTOR protein.

So how does the algorithm find the local optim within proteins?

With autodock, a population of individuals (ligands) are randomly placed within the receptor. The conformation ligand-protein interactions are measured for each individual and is then followed by a conformational "mutation" (See image below).

Ligand "mutation".

The binding energy for each conformation "mutation" is measured until a local optimum for a specific population of individuals is reached. The binding energy of the local optimum of each population is measured, and the global optimum is the population of individuals that have the best binding energy (See results below).

If the evolutionary algorithm is well designed, the conformation of the global optima will correspond to the experimentally determined crystallographic pose. The Root Means Squared Deviation (RMSD) of a docked ligand compared the to the crystallographic pose is generally used as a good indicator. A RMSD value less than 2 is considered a success. In the case of the Autodock software, the global optima is supposed to correlate with the crystallographic pose (RMSD <2). As an example, a ligand was docked into a protein with the following results.

Docked ligand positions and binding energies

Now let's consider another example in nature and how heritability and selection is applied.

As an example, consider the following diagram:

A fitness landscape (From here)

Again, a few important aspects from the figure:

• 1) Fitness depends on the phenotype
• 2) Fitness in this case is the capability of the phenotype to reproduce (self-replicate)
• 3) The parameters for succesful self-replication are many. A few examples:
• A) Fast replicators (e.g. bacteria)
• B) Intelligent replicators (e.g. monkeys)
• C) Cooperative replicators (e.g. ants)
• D) A combination of the above (e.g. humans)
• E) Population dynamics
• F) And others...etc.

Therefore, if certain parameters are not on a fitness landscape for a certain phenotype, such as the capacity to construct a car, such a trait will not be selected in the next generation if the population of phenotypes consist of bacteria)

One thing that is interesting about the docking software is that because it seeds the ligands randomly within the protein and the position of the protein is "mutated" randomly, you will get different results every time. However, docking runs still converge on a the same global optimum after the evolutionary algorithms were completed. And the global optimum corresponds reasonably well to the crystallographic pose (the optimal design). That happens if the software is well designed of course.

There are many parallels between this evolutionary algorithm and the history of our universe. These include:

1. Incorporation of randomness (with regards to fitness) as well as selection.
2. The process is biased towards a few ends just like our own evolutionary history (e.g. An End to Endless Forms: Epistasis, Phenotype Distribution Bias, and Nonuniform Evolution)
3. Convergence. Our evolutionary history is filled with examples of comvergence. E.g.:

• The spectacular convergence of abiogenesis into a universal highly optimized genetic code that governs just about all life forms on earth.
• Beautiful structural convergence on several levels. e.g. Convergent Evolution
• Molecular convergence: Carbonic anhydrases, Prestin, Others

If we are indeed living in a simulation, an evolutionary algorithm of some sort seems plausible.

Question is:
Is this conception of reality with information and algorithms being fundamental categories really compatible with the mechanistic/anti-teleological conception of the material world?

Nano-intentionality and Molecular Autonomous Agents

Intrinsic intentionality and inherent goal-directedness of eukaryotic cells is defended by Tecumseh Fitch and minimal molecular autonomous agents are characterized in "On emergence, agency, and organization" (by Stuart Kauffman and Philip Clayton).

The "aboutness" and "goal-directedness" of eukaryotic cells and how it relates to nano-intentionality is defined as follows (p14):

The crucial pre-mental properties of a cell are that it can
1) respond to (somewhat) novel circumstances, eventualities for which it is not specifically-prepared by the evolutionary "memory" instantiated in its DNA.
2) discover, through an individual process of trial and error, some "adaptive" (in the physiological sense) response or solution.
3) in various ways incorporate the results of this discovery into its own structure, thus "recording" or "remembering" (in a non-mental sense) this past, individual history.

It is argued that simple single-celled eukaryotes possess nano-intentionality and it is stressed that one of the abilities of a nano-intentional structure is its ability to rearrange its physical structure in response to environmental circumstances. An example of eukaryotic chemotaxis (sensory adaptation) in the amoeba and its ability to react to environmental signals and adapt to them by inducing structural changes was given, e.g. when seeking and ingesting food was given. Chemotaxis involves structural changes in response to environmental circumstances and it is not limited to eukaryotes as bacterial cells are also capable of chemotaxis. In this respect, bacterial cells would qualify since no other reason was provided for not including bacterial cells.

Kauffman and Clayton argue that the simple example of a bacteria that is able to swim up a glucose gradient is an example of an organism acting on its own behalf and they call such a system a "molecular autonomous agent". They continue to provide a tentative five part definition of a minimal molecular autonomous agents (p505):

Such a system should be able to
1) Reproduce with heritable variation.
2) Perform at least one work cycle.
3) Have boundaries such that it can be individuated naturally.
4) Engage in self-propagating work and constraint construction.
5) Be able to choose between at least two alternatives.

The earliest life forms emerged about 3000-3400 million years ago (ref) and were likely bacteria.

Many natural sciences aim to detect the actions and intentions of agents. A few of these include forensic science, archeology and SETI.

If it is accepted that bacteria qualify as nano-intentional molecular autonomous agents then it would seem evolutionary and molecular biology also falls squarely into the category of natural sciences concerned with detecting the actions and intentions and/or "nano-intentions" of agents over time, be it mental or pre-mental.

If ID science is defined as "detecting the actions and intentions and/or "nano-intentions" of agents over time" does this mean evolutionary and molecular biology qualify as ID science?