Biological Info: Understanding & Analyzing

[BillTre]

TL;DR Summary

Biological Information: how should it be conceived of?

"Biological Information" is a common concept in some parts of biology. It most often (in my experience) is used to refer to the sequence information in nucleic acids (nucleotide sequences in RNA and DNA) and in proteins (amino acid sequences in proteins). This is most apparent in the flow of information found in the "central dogma" of molecular biology (DNA --> RNA --> Proteins).

This seems, to me, to be a rather limited view of the information content of biological entities.
There is obvious information content in the many different molecules found in even the simplest organism. Examples of this might be what kinds of proteins are found in what quantities, at locations, at different times, in a cell.
There are lots of other examples.

Sometimes a distinction made between "analog" and "digital" (sequence) information in biology, in an effort to acknowledge the (ill-defined) non-sequence information (without really giving it serious consideration) found in the other (non-sequence) stuff of cells.

I would prefer a more bottom-up approach from molecules to larger molecular assemblies and their function and the information involved in their functional interactions, but have not seen this. This, in someway, might be built up into the really complex functional molecular machinery of the cell.
I have not seen this in any of the literature or information analyses that I have seem (dealing with DNA or RNA counts). Efforts counting up the information content in cells has not seemed very impressive to me.

Other related issues are:

• The numerous copies of a given protein (or other molecule) present in a cell could be doing different things at the same time, like a very complex parallel processor.
• If one considers the relations of DNA, RNA, and protein from a Shannon information point of view (for example if mRNA is a tape of production commands), additional protein and RNA components provide the required interpretative function underlying the "central dogma" flow of information. The interpretive functions in this case are the transcription and translation mechanisms that are universal in contemporaneous biology.

I am wondering (and soliciting the opinions from people here about):

• What might be a more generally applicable concept of "biological information"?
• As atoms are combined into molecules, and molecules are combined into larger functional entities, in what way is information construed as arising from these new assemblies/interactions?
• I am aware that in quantum mechanics, there is information associated with atoms (and, I assume, atoms assembled into molecules). Is this information based on the quantum composition (in properties and sub-sized components) and/or their wave functions? It seems to me that building these concepts up to the scale of functional molecular systems could be a possible way to generate "biological information", but that due to the complexity of these situations, they would be computationally out of reach.

At this point I am only interested in conceptually, not quantitatively, thinking about these issues.
Quantitative analysis of biological information content could well be computationally out of reach due to the large numbers of independently interacting molecules.

 

[Fra]

Just a reflection. I think the quest of an ultimate - objective - eternally true, information measure, is misguided. I think this is true both in physics as well as biology.

additional protein and RNA components provide the required interpretative function underlying the "central dogma" flow of information. The interpretive functions in this case are the transcription and translation mechanisms that are universal in contemporaneous biology.

I agree this is a key point. One can not isolate the code, from it's interpretative and supporting context. Not in biology and not even in physics. In physics the information implicity in the "laws themselve" do not count. This is what layes the ground fore all the pathological fine tuning problems, that in biology are solved by evolution. And I think the evolution is inherently relational, it is not something going on against an objective backdrop like cellular automata.

It seems to me that building these concepts up to the scale of functional molecular systems could be a possible way to generate "biological information", but that due to the complexity of these situations, they would be computationally out of reach.

.. those attempts typically end up with computability and fine tuning issues. This is pathological i think both conceptually as well as evolutionary. Biological systems need regulatory strategies that are working. Ie. that can be executed in real time whilst trying to survive and reproduce in hostile environments. Evolution simply will put systems out of business that fail to be responsive. I think the biologist understand this naturally much better than physicists. This is the failure of reductionism the dominates still in physics.

I think physicits have more to learn from evolutionary biologists, than vice versa. This is one of the reasons I am interested in biology and cellular regulation from evolutionary perspective even from perspective of foundation of laws of physics.

/Fredrik

 

[*now*]

Just a reflection. I think the quest of an ultimate - objective - eternally true, information measure, is misguided. I think this is true both in physics as well as biology.I agree this is a key point. One can not isolate the code, from it's interpretative and supporting context. Not in biology and not even in physics. In physics the information implicity in the "laws themselve" do not count. This is what layes the ground fore all the pathological fine tuning problems, that in biology are solved by evolution. And I think the evolution is inherently relational...

From this is it your suggestion here that the laws-themselves do count relationally?

 

[Fra]

From this is it your suggestion here that the laws-themselves do count relationally?

This can be misinterpreted, but to comment briefly for a change i would say yes.

What i mean is that "information about" the initial conditions should be on equal footing as information about laws. The assymmetry we have today in physics is confusing if you think of laws as a result of evolution. Evolution then does NOT mean "dynamics".

/Fredrik

 

[BillTre]

A big difference between considerations of information in biology and non-biology is the large impact the context in which an informational thing (like a molecule) finds its self.

• in a cell a mRNA can have a big distinctive impact due to the products of transcription it can direct the production of in a cell
• when not in a cell, it is just a big molecule floating around waiting to be degraded (unless it gets into a cell).

From cell to cell, differences in the cell's internal context can affect the molecules it contains.
The functional importance of the information of a particular molecule depends, not only the context that affects what it can do (like what a mRNA can produce), but it also depends on the context within which its effects play out in (a protein in cell A may have an effect by binding a molecule which is not present in cell B).

 

[Fra]

This contextuality should be more acknowledged also outside biology, this was my point. The traditional solution is to just find a sufficiently detailed view, where the context itself is also modeled and then the "relations can be described" from an objective stance, but this is I think missing a deep point, which is that no agent can hold infinite information and even less process it in a viable manner in realtime unless it has infinite time. This means that evolutionary learning, implies forgetting as well. A kind of lossy compression, but the optimal compression is context dependent. Ie. what skills or traits that are less important depends on the environment, which is also subject to evolution. And when you try to understand the actions or "logic" of a system (agent or life form) then the logic exists only relative to it's own history. What is "true" in one context, can be "false" in another.

Edit: IMO, this is why current fundamental physics and QFT makes the most sense only in it's asymptotic predictions, scattering etc. Which in turn, makes real sense only for small subsystems, ie. atoms, from the perspective of a big lab. So the reductionism do work, but only in a certain domain, after that it's explanatory value fades. In particle phycists this limiation is not obvious! In biology it is obvious. But at least in from some ressearch programs, the limitations become visible also in physics as you try to unify forces WITHOUT getting into finetuning problems, or renormalisation problems.

/Fredrik

 

[Jarvis323]

It's kind of interesting to me to think about life in the context of information theory and entropy. It's really complicated.

Life is an example of order from chaos, and order means less entropy and less Shannon information. But it's pretty unsatisfying to think about life as being low in information content compared to something like plasma, and to think of ordered and patterned things as those which have less information than something random and disordered.

At the same time, disorder from chaos and randomness helps a system to explore a larger state space more uniformly. And that gives the system a chance to find itself in one of those states where the future trajectories are more patterned. And then the system can be stuck in those regimes for a while, until the trajectories fall out of the regimes and the systems become chaotic again.

But all of this thinking is really questionable I think, because as you have been stressing we are talking about open systems where the environment is evolving alongside the object inside of it. We're a part of the environment and the environment is a part of us. And as far as we can be concerned, the environment is infinite, and chaotic, and only partially measurable, and so these trajectories we're talking about are actually bigger and more complicated than I can comprehend, and I'm not sure if the concepts applied when talking about the closed system are the same when talking about the open system.

 

[BillTre]

Life is an example of order from chaos, and order means less entropy and less Shannon information. But it's pretty unsatisfying to think about life as being low in information content compared to something like plasma, and to think of ordered and patterned things as those which have less information than something random and disordered.

I think of life more as being highly structured rather than as just a particularly big pile of Shannon type information. But an important part of that structure involves how information (and the material carrying it) works in the living entity.

I would also say its propagated order, rather than order from chaos.
One might consider life's 'origin" as order from chaos, but it probably can from special, fairly well ordered, environments, with a lot of available energy to be made use of. I would call these nursery environments. Not chaotic messes.

 

[Jarvis323]

I think of life more as being highly structured rather than as just a particularly big pile of Shannon type information. But an important part of that structure involves how information (and the material carrying it) works in the living entity.

I would also say its propagated order, rather than order from chaos.
One might consider life's 'origin" as order from chaos, but it probably can from special, fairly well ordered, environments, with a lot of available energy to be made use of. I would call these nursery environments. Not chaotic messes.

The idea is that without chaos those special environments might never arise. They are like jackpots on a slot machine and chaos spins the wheel, so that eventually every configuration shows up. Without that, a universe of non-nursury environments could happilly remain non-nursury environments forever.

 

[BillTre]

I like the chaotic aspect for creating the diversity of mixtures of chemicals initially brought together as potential units of life, from with in the diversity of nursery environments. The most often considered environment for biological origins are now alkaline hydrothermal vents.
The small differences in the vent environment can affect which chemicals are present at each location.

The general environment itself can easily be generated by the contact of ocean water with malfic or super-malfic basalt. The resulting chemicial process is called serpentinization (more details here). That particular kind of basalt is generated from solidifying mantle. There was a lot of that on the early earth, wwhere water was in contact with the new ocean floor. At some sites, these reactions can go on for thousands (or more) years, thereby providing incipient life with a nice, long lived, nursery site.
The sepentinization process creates H₂ and CH₄, which in turn though interactions with natural catalysts (Fe and S and sometime Ni, also found in the output of the alkaline hydrothermal vents), can form a variety of molecules currently found in core metabolism, as well as molecules similar to membrane molecules. This stuff needed to generate life.

Fischer-Tropsch-type reactions are unique among abiotic organic processes in providing a source of linear molecules. This is important in a prebiotic context since, for instance, linear fatty acids are essential for the formation of the bilayer membranes

Molecules (like lipids), which can make membranes (bilayer lipid membranes). Membranes generated could then package up small (and varying) samples of chemicals, which may or may not be capable of self-replication.
These catalysts are often, today, found in the catalytic centers of protein enzymes.

Serpentinization has also been proposed as a potential source of the intermittent production of methane observed on Mars.

Due to the particular situations of the local geology, the ionic composition can vary from place to place. This, in my point of view, would be one of the chaotic aspects of this process. The process itself was going to occur in the Earth's past, and is occurring now.
Its also expected to occur on particular kinds of rocky planets throughout the universe.

Now, such alkaline hydrothermal vents are restricted to areas (a few km) on either side of seafloor spreading places, or where chemically appropriate volcanic lava has solidified and exposed to water.

A similarly proposed environment for origin of life is transitional terrestrial ponds found on the sides of volcanoes. They have similar chemical reactions as well as the possibility for wet-dry cycling, which can polymerize amino acids and nucleotides (to make peptides and nucleic acids). These can also be observed today and have chemical differences between different sites.

The much hotter "black smoker" hydrothermal vents, are much closer to the much hotter magma and recently solidified materials. They are chemically different.

 

[Jarvis323]

There is so much unknown I think about how life forms, and how rare and widespread it is. My thoughts about it regarding chaos and order and information are pretty general and vaugue. Even if life does form, it still has to last, and then it has to evolve to survive. Who knows how many planets there are where life formed and then almost immediately went extinct. Because of chaos, we can predict that somewhere out there, eventually, the conditions will be just right.

And then when it comes to survival, life also relies on chaos/randomness to generate diversity and ensure that between the changing environment and a subset of the life forms, conditions will continue to be right for life to continue on. And then because the environment varies so much, enough different kinds of life emerge out there that as the environment keeps changing, there should be likely some life somewhere that is suited for the next conditions and that should last for a while. So I think there is an important interplay between randomness and order when it comes to life. Life sort of catches and then rides a wave of chaos.

People tend to think of randomness as void of patterns. But true randomness really means you'll find every pattern somewhere in there.

Since information theory measures the randomness in a process, it seems it should be a good framework for studying the relationship between randomness and order and life should be a perfect example. And lots of people try this, but it's really complicated.

 

[Fra]

People tend to think of randomness as void of patterns. But true randomness really means you'll find every pattern somewhere in there.

For me a key thing is to note that we always have what I think of as "interacting perspecives". Ie. what is void of patterns, and what does hold patterns is partially in the eye of the beholder. And the decoding capacity of the observer. This distinction is important at least in the "agent interaction" interpretation, as the action of agent towards the environment, depends on what patterns it can or can not decode in real time. Ie. a fast decoder will have different observable behaviour, even as seen from an external observer.

Most laws of physics are focuses on laws as "constraints". This makes sense for closed systems. But for open systems I think it makes more sense to think of the laws as guiding rules of selfpreservation which is the key to understand the rules for self-organisation in open systems. And interactions are part of evolution, can just can not be described as dynamics following eternal timeless laws.

/Fredrik

 

[Laroxe]

This is a fascinating issue, and the linear way in which people have understood how biological systems work have and continue to hold back progress in understanding these systems. We continue to see papers identifying specific genes with biological effects, which rarely happens and usually based on pathology.

Information is often defined as some sort of stimuli that has a meaning, in a particular context, for its receiver. This means that the receiver must be aware of the stimulus, recognize it as containing information then, depending on the current state of the receiver, extracting some sort of meaning. In fact, each part of this sequence is dependent on other information, and in each case the sequence can result in highly variable effects. This means that a stimulus following exactly the same linear sequence, with the same molecular changes, may not result in predictable effects on the receiver. The current attempts to understand biological information transfer is much more embedded in network models, so the actions of genes can only be understood in terms of the actions of the networks of other genes that impact on its activity. Each of these genes being subject to their own control environment and feedback from the intracellular environment and the RNA that their activity transcribed. This is then compounded by the fact that virtually all physiological processes can impact on gene expression, just as gene expression can impact on most aspects of physiology.I think one problem is in fact identifying a model that can effectively be generalized as an explanation for biological information. While evolution is dependent on genetic changes, we are trying to make sense of the end product of the evolutionary process, and that is the whole organism and how it functions in its environment. This is hugely complicated and information works in all levels of the system, really, its only relatively recently that we have started to get some sort of insight into these processes. I think currently, there are far too many gaps and uncertainties to think we are anywhere near making sense of it all and it being biology I doubt that we will ever get beyond predictive probabilities.

 

[Fra]

My own personal journey started in chemistry, in particular it's dynamics and thermodynamics, like what determines the rate of reactions etc, and laws of chemistry is supposedly explained by physics, and the laws of physics, so that is where i went looking for the answers. Then years later I spent some time on understanding yeast in the context of brewing science, entertaining the idea of a first principle simulation of a beer fermentation, but found then analogies between human decision making, cell regulation logic, and the interaction rules of an elementary particle seemed to have a lot in common. These insights for me drastically popped the illusion of "reductionsm" that everything is ultimately explaince in terms of the laws of physics, and all else is just an "application" - although a complex one. That thinking is deeply wrong, and I had been confused for a number of years, and paradoxally it's becoming acute again during unification of the laws of physics. Alot indicates IMHO that (see also The Evolution of the Laws of Physics - Lee Smolin (SETI Talks)) also the laws of physics are subject to evolution. Smolin even argues that a lot of problems in physics (unifying interactions) steem from a lack of this insight. (That said, nonone has the new theory yet, but the problem is that some people insist on the old paradigm, Smolins agenda has partly been to open the mind of physicits that as are stuck in traditional thinking. But I think that most are still stuck, and they consider these ideas nutty).

/Fredrik

 

[Buzz Bloom]

I can conceive of a planet (Earth) in its early time having many different chemistry phenomena leading to many different pre-life systems. I can conceive that of the many different pre-life forms, only one was much more competitive than all the others, so only one survived as life chemistry as we know it today. I can conceive that some other planet somewhere in the Milky Way may have had a somewhat different early pre-life chemistry environment which may have led to a biochemistry different from Earth's, and possibly more than just one (maybe two) surviving independent chemistries of life.

I am unable to guess (even roughly) at some specific time in the future when research on Earth might discover some distant planet with life on it with a very different life chemistry that that on earth.

 

[jim mcnamara]

Some few links on the too many subjects in this thread for a quality thread, IMO.
Some links are speculative subjects.

Proteasomes -
Proc Jpn Acad Ser B Phys Biol Sci v.85(1); 2009 Jan
Microbiome project -
https://en.wikipedia.org/wiki/Human_Microbiome_Project
Epigenetics -
https://en.wikipedia.org/wiki/Epigenetics
RNA world
https://en.wikipedia.org/wiki/RNA_world
abiogenesis
https://en.wikipedia.org/wiki/Abiogenesis