Menu directory status & updates copyrights help

MindCode: DNA transcription to RNA, information process rather than protein build

Table of Contents


Preface

05Nov2023 Everything on this [DRAFT, incomplete, error-filled] webPage is comprised of [question, "what if?" speculation]s that are limited by my lack of [cellular, genetic] knowledge. The idea is to go through and dream up what I can, then look at great ideas in [research papers, biology] to see may own [gap, limitation]s. I haven't even started to gather together [past, present, future] references.

This webPage does NOT provide solid [hypothesis, theory], but instead presents "what-ifs" in a "multiple conflicting hypothesis" context. I am not concerned with [right, wrong, true, false] at this stage, only in generating questions, rather than answers, that may somehow prompt diverse ideas, and not to become a tool of [concept, theory], rather than the other way around.

The ONLY MindCode aspect that concerns me at present (07Nov2023) is to produce computer programs related to the section below "Computations with multiple RNA strands". Computer programs formalize thinking in a seerious way, reveal the blemishes, and suggest possible avenues for [progress, failure], which are nearly impossible to distinguish.

I am sure that I have no new ideas to contribute, that everything here has been thought through by others, many times in many diverse ways. See the section "Shapiro & Benenson: example stream from early MindCode on..." below for a great example of my state of ignorance, and the beauty of serious work. After my initial speculations, I will look much more closely for diverse approaches in pre-existing research, and for hints from biological experimentation not reflected in mainstream [model, conclusion]s. For that I have to rely on others who can pick out the [gap, weakness, contradiction]s . That is more important to me than simply picking out the "accepted mainstream concept".

For a decade or two, I was not expecting to proceed with MindCode, as I felt that "professional researchers" would be doing something similar. That has happened, but no-where nesar to the extent that I was expecting. Furthermore, for my own learning, this subject is a priority for me, and doing something (computer programming) is a primary way for me to feel that I am learning in a non-trivial way.


[Turing, von Neuman] machines of gene mechanisms?

The current paradigm in Artificial Neural Networks (ANN) research assumes that neurons fire when their membrane potentials exceed a [flexible, adaptive] level. However, I break with that tradition by a "what if" assumption that firing also occurs on the basis of information processing "programs" in each neuron, which : Traditional Artificial Neural Network research does NOT go into genetics, but there are ANN areas that do address that subject. It seems strange to me that researchers of Spiking Neural Networks (SNNs) claim biological plausibility without any genetic basis. Granted, their models don't require that level of detail, and in any case perhaps that kind of information is problematic.
.
Probably the most important section of this webPage is "Computations with multiple RNA strands". Most other sections provide context.

You can see that I am closely following normal [von Neuman computer, simple modern computing language] concepts. In earlier versions of MindCode, I even called many operators "Z80-type" because I didn't have [time, knowledge] to try to build biologically-based operators. I certainly do not believe that biology is limited to that context, but I probably am. Others have probably tackled this and come up with many ideas.

Finally (for now) I should mention "Super-Turing" machines that cannot be described by Turing's theories. My very scant understanding is that evolutionary processes are one key for this. Hava Sieglemann really hammered this into me.

[intra, extra]-cellular processes, [neuron, astrocyte]s

Stephen Grossberg's 2023 "Conscious Mind, Resonant Brain" book touches on the subjects of [paleontology, evolution, organisms], albeit with minimal reference to how gene mechanisms work in neurons and the brain. That book is a foundation for my long-term Neural Networks research goals, so I prefer to introduce my current MindCode project on [Turing, von Neuman]-like machine potential by extending Stephen Grossberg's description of [intra, extra]-cellular neural processes. Just for fun, and prevent myself from forgetting about astrocytes, I have included a sub-section towards the end of the is webPage.


Cellular mechanisms for [protein, information]

Can we use the SAME cellular gene [structure, mechanism]s for information processing, as are used for protein contruction? As it would be economic to do so, I will work under that assumption except, and which that cannot be accommodated by simple adjustements to common [structure, mechanism]s.

[DNA, rhibosome, etc] addresses

06Nov2023 very important subject, for later... some challenges : My guess is that biology has long done this, but perhaps with very interesting twists, and in so many ways.

[2, 4]-value logic for [protein, information]

Does the direction of an acid-base pair matter (with respect to a specific strand of the double-strand DNA helix)? Even if that is NOT the case for protein construction, it may still matter for more abstact information processing (see 4-value logic below). 4-value logic is NOT typically used in formal logic, and I have no idea of whether it is used biologically. But I retain it anyways at an early stage to keep it in mind, and in any case it can always be reduced to Boolean logic. Do the cells use it for : At any rate, if biology uses 4-value logic, it certainly "understands it" far better than I do.

ribosomes

Because ribosomes already serve to build proteins from enzymes on the basis of RNA codons, they are a natural place to look for RNA program execution as well, perhaps in combination with protein processing. If that is the case, then :

microtubules: platforms for [transport, info processing]?

Many papers have implicated microtubules in the transport of [material, RNA?] for neurons. But how do they fit in with genetic machinery and its component mechanisms?

same mRNA code, different mechanism: ergo diff [program, protein]?

11Nov2023 Everybody might wonder if the same mRNA code might be interpreted differently by [local, specialised] mechanisms. That is particularly the case for "code as a program", as that is regularly a feature of local [data, class] symbols in computer programming languages, where one wishes to avoid "cross-over effects acros operators. Bash scripting does NOT have this feature, so I use unique 3-digit number codes to end each symbol in libraries to give something like that isolation. This is not a complete solution, but is [simple, pragmatic, normally effective].

It's easy to wonder if that also applies to protein-building mRNA strands. Hopefully, there is well-establish data on this in the protein-building research area...

missing sub-sections for genetic machinery

05Nov2023 I am missing many sub-sections related to detailed genetic machinery to explain how it works.

DNA transcription to mRNA

13Dec2023 (https://en.wikipedia.org/wiki/Transfer_RNA)
"... Transfer RNA (abbreviated tRNA and formerly referred to as sRNA, for soluble RNA[1]) is an adaptor molecule composed of RNA, typically 76 to 90 nucleotides in length (in eukaryotes),[2] that serves as the physical link between the mRNA and the amino acid sequence of proteins. Transfer RNA (tRNA) does this by carrying an amino acid to the protein synthesizing machinery of a cell called the ribosome. Complementation of a 3-nucleotide codon in a messenger RNA (mRNA) by a 3-nucleotide anticodon of the tRNA results in protein synthesis based on the mRNA code. As such, tRNAs are a necessary component of translation, the biological synthesis of new proteins in accordance with the genetic code. ..."




Computations with multiple RNA strands


A simple-to-harder progression of sub-sections provides examples that the reader can [modify, improve] themselves.

The general concept is that a "functionally-specialized chef" moves back across strands of input RNA, producing output RNA strand(s) based only on the input RNA coding. I re-emphasize here that we are considering program [command, data, control, etc] for INFORMATION transformation, rather than codon sequences for building [enzyme, protein]s. I don't know how far the "information versus [enzyme, protein] metaphor" can be pushed.

Where-ever the metaphor leads us, it would be much simpler if the same [enzyme, protein] could be directly applied to programs, so I will see how far that gets me. Yes, I am lazy.

Very [incomplete, initial] lists of example [function, operator, procedure]s that I have been considering appear on my very incomplete program webPages : I am keeping the commentary, program]s separate to avoid the usual problem of [handle, document]ing strong [concept, coding] changes. That also provides a long-term record that I may use with current work. Note that the "MindCode" term arose ~1992-2002?. I then made it my license plate number to keep reminding me that I was always diverted into other projects, and rarely worked on MindCode. I've kept up that tradition ever since.

I have yet to eleaborate on mechanisms of genetic machinery to do this.

NOT: simple function of a single RNA input strand, one output RNA

boolean NOT transforms [ATGC(U)] sequences to, let's say arbitraril: [A,G] sequences, with A being zero (false), and G being 1 (true). The "chef" therefore may have an "inventory" of [AG] [acid, base] amines, and I have yet to think though whether it should be an acid for 0 and base for 1, or if they should have the same OR[acid, base] nature.

4-value logic NOT Uh-oh! It's easy to transform a certain [true, false] to its NOT, but is that actually correct? What about the 4-value [mostly-correct, unknown] states? I will soon be working on that. Hopefully, I can use the


AND: simple function of two input RNA strands, one output RNA

boolean AND

4-value logic AND

GoTo

I can hear everybody groaning at the sight of an old nemisis of modern computing: "GoTo". Too bad I can't hear your groans, but the Basic programming language showed me that it has its own beauty and power <grin>.

Given my primitive knowledge of genetics, GoTo seems to be a nice speculative match for 5' start sequences, both [protein, program]. Typically, 5' starts simply search for matches of the subsquent codon sequences in the DNA. But not so long ago most descriptions didn't include a role for micro-tubules either.

Does biology have "[relative, absolute] addressing" (relative - local proximity on same DNA or RNA strand, absolute - address may even be on different chromosome)? I don't remember any references mentioning that possibility. In a previous sub-section of this webPage, I have provided a few (incomplete) points on addressing. I have a lot of [read, program]ing to do here

IF: control branching to an [operation, address]



???



MindCode [learn, evolve]: Grossberg's 'Conscious Mind, Resonant Brain'

While I have long been a fan of the work of Stephen Grossberg and his colleagues, I was very surprised with his 2021 Book "Conscious Mind, Resonant Brain". (This link shows a menu that lists details of many themes from his book, plus commentary on well-known concepts of consciousness.) His book went far beyond my awareness of his work (obviously I was horribly out of date). [Right, wrong, true, false], it also goes far beyond any other [concept, work] that I am aware of in explaining how [neurons, the brain] work. The results are not simple, nor are they amenable to the normal "yap, wave your arms" that we all like so much. Maybe that's why it's not so [popular, well known]. To me, concepts of consciousness that do not [emerge from, work with] non-conscious processes, and which do not ellucidate mechanisms, are not satisfying, even though they can still be pragmatically useful.

In any case, for now Grossberg's work provides an experiment-iteration-based, rational concept platform that I feel comfortable with for extending MindCode and my other neural network projects.

This contrasts to more popular (and very successful), [statistical, information theoretic, modern existential] science basis. Don't get me wrong, I very much like statistics, and especially information theoretics (kind of checmical engineering themodynamics for thinking?). One fun example was Harold Szu's information-theoretic derivation o Hebb's law of learning. But in the end, statistics can help point the direction to mechanistic understanding, but rarely gets you there. It seems to me that "theoretical understanding" of mechanisms may often be "after-the-fact" based on practitioners rather than theorists. Grossberg is an exception in doing both.

In any case, I will work on Grossberg's concepts in the future, and apart from providing a simple "figure-captions-based" thematic overview of his work, my only other comment is on the subject of consciousness (below).

Consciousness: Grossberg's tie-in of [, non] conscious processes

There are only two concepts of consciousness with which I am comfortable, biologically based concepts from Grossberg and colleagues, and the late John Taylor's "advanced control theory" concepts for consciousness (linked webPage not built yet 07Nov2023). But the latter is not amenable at the present time with the directions of MindCode, with a special emphasis on genetics. I did do a very [quick, incomplete] commentary on consciousness concepts, and a simple overview of [definitions, models] of consciousness.

To me, consciousness iteself is interesting, but it is an "easy problem" (sort of) in the context of Grossberg's work. The MUCH harder problems are Grossberg's non-conscious processes, and, somewhat less difficult, his tie-ins of non-conscious with conscious processes, the latter being emergent from the former.


Biological context

[pro, eu]karyotes

The following lists are reminders to myself.
Cell micro-structures Cell components and organelles : gel-phase water (EZ-water) - the fourth phase of water? Gerald Pollack's concepts are not mainstream, to say the least. But he does point out gaps in mainstream thinking, and [right, wrong, true, falwse] his gel phases of water may help. Here are a few items from his bag :

Lamarckian versus Mendellian heredity, spiking MindCode as special case

I'm gone over this before, and it is fascinating to me. I am keeping this section as a "placeholder" to return to in the future...

This may also be relevant to the issue that inter-cellualr processing via spiking and possibly other mechanisms may still be analysed on the basis of underlying program coding, not just protein coding?

Astrocyctes

Here are some [wild, crazy, radom] speculations of what astrocyctes are doing, if for no other reason that they have long intrigued me, and I don't want to forget about them. It's made me realize that 2 decades have passed since I last did a tiny bit of reading. Much must have happened since then... There are apparently 10 times as many astrocytes as neurons, so is it possible that, among other things, they are a major location for [many, most] intra-cellular computations? silent, massively parallel? If so, how do they signal neurons?

If I remember correctly, complexity analysis (information theoretic? I forget) have long commented that maximum information [storage, processing, complexity] occurs with a mix of [local, long-range] connections. Olaf Sporn's "rich club" of neuron centers may act in the same way at higher levels of brain organisation?

Ontogeny: the growth of bioFirmWare

bioFirmWare = [architecture, function, process, operating system, model, control, plan]
all "categories" are fuzzy, rather than crisp?
...includes body, brain, etc, etc...

For later commentary (from old notes too).


Biological similies

Endocrine system

This an area that is fascinating to me. I don't know how many years that it may take for me to get to work on it. But like much of the rest of physiology, I can't help thinking that advanced control theory concepts might play a role. Stomach digestion of food is another example, and one that I certainly have no real conscious knowledge of, but it gets done.

bone [shell, fibre]s

Building MindCode from [bio, psycho]logical data

[proteomic, neuroinfo, MindCode]*[protein, program]*bioFirmWare

bioFirmWare = [architecture, function, process, operating system, model, control, plan, etc]
all "categories" are fuzzy, rather than crisp?
...includes body, brain, etc, etc...

The interaction of "fuzzy" categories, as listed in this section's title, offers a potential starting point, experimental large-scale data, and check on each of the categories.

MindCode [identify, library, model, predict, change, heal]

Without referring to my earlier [note, document]s, at this time I will restate what I remember of some of the longer-term objectives of MindCode, with many extensions :
Of [principle, practice]s :

MindCode applications to keep in mind

This section is definitely "arm-waving, yapping" on my part. Not soemthing that I am fond of unless it helps to inspire other work for [myself, others] like science fisction long has.

Autonomous systems, networks, robots

With the hyperebolic excitement over the "new AI", specifically Transformer Neural Networks and Large Language Modelels like [Google Lambda, BARD], OpenAI chatGPT, llama, etc], it is natural to address this subject. But I will wait for pehaps a year or two, when I have sufficiently progressed with the biological side, which is still far more interesting to me.

Missing concepts - among hundreds



Shapiro & Benenson: example stream from early MindCode 2005 on...

+-----+
https://www.nature.com/articles/35106533
nature letters
Published: 22 November 2001
Programmable and autonomous computing machine made of biomolecules
Yaakov Benenson, Tamar Paz-Elizur, Rivka Adar, Ehud Keinan, Zvi Livneh & Ehud Shapiro
Nature volume 414, pages 430–434 (2001)Cite this article

Abstract
Devices that convert information from one form into another according to a definite procedure are known as automata. One such hypothetical device is the universal [Turing, von Neuman] machine1, which stimulated work leading to the development of modern computers. The Turing machine and its special cases2, including finite automata3, operate by scanning a data tape, whose striking analogy to information-encoding biopolymers inspired several designs for molecular DNA computers4,5,6,7,8. Laboratory-scale computing using DNA and human-assisted protocols has been demonstrated9,10,11,12,13,14,15, but the realization of computing devices operating autonomously on the molecular scale remains rare16,17,18,19,20. Here we describe a programmable finite automaton comprising DNA and DNA-manipulating enzymes that solves computational problems autonomously. The automaton's hardware consists of a restriction nuclease and ligase, the software and input are encoded by double-stranded DNA, and programming amounts to choosing appropriate software molecules. Upon mixing solutions containing these components, the automaton processes the input molecule via a cascade of restriction, hybridization and ligation cycles, producing a detectable output molecule that encodes the automaton's final state, and thus the computational result. In our implementation 1012 automata sharing the same software run independently and in parallel on inputs (which could, in principle, be distinct) in 120 μl solution at room temperature at a combined rate of 109 transitions per second with a transition fidelity greater than 99.8%, consuming less than 10-10 W.

+-----+

nature letters article
Published: 28 April 2004
An autonomous molecular computer for logical control of gene expression
Yaakov Benenson, Binyamin Gil, Uri Ben-Dor, Rivka Adar & Ehud Shapiro
Nature volume 429, pages 423–429 (2004)Cite this article

Abstract
Early biomolecular computer research focused on laboratory-scale, human-operated computers for complex computational problems1,2,3,4,5,6,7. Recently, simple molecular-scale autonomous programmable computers were demonstrated8,9,10,11,12,13,14,15 allowing both input and output information to be in molecular form. Such computers, using biological molecules as input data and biologically active molecules as outputs, could produce a system for ‘logical’ control of biological processes. Here we describe an autonomous biomolecular computer that, at least in vitro, logically analyses the levels of messenger RNA species, and in response produces a molecule capable of affecting levels of gene expression. The computer operates at a concentration of close to a trillion computers per microlitre and consists of three programmable modules: a computation module, that is, a stochastic molecular automaton12,13,14,15,16,17; an input module, by which specific mRNA levels or point mutations regulate software molecule concentrations, and hence automaton transition probabilities; and an output module, capable of controlled release of a short single-stranded DNA molecule. This approach might be applied in vivo to biochemical sensing, genetic engineering and even medical diagnosis and treatment. As a proof of principle we programmed the computer to identify and analyse mRNA of disease-related genes18,19,20,21,22 associated with models of small-cell lung cancer and prostate cancer, and to produce a single-stranded DNA molecule modelled after an anticancer drug.

+-----+
https://www.sciencedirect.com/science/article/abs/pii/S0022519321004045
Journal of Theoretical Biology
Volume 537, 21 March 2022, 110984
Journal of Theoretical Biology
An RNA-based theory of natural universal computation
https://doi.org/10.1016/j.jtbi.2021.110984

Abstract
Life is confronted with computation problems in a variety of domains including animal behavior, single-cell behavior, and embryonic development. Yet we currently do not know of a naturally existing biological system that is capable of universal computation, i.e., Turing-equivalent in scope. Generic finite-dimensional dynamical systems (which encompass most models of neural networks, intracellular signaling cascades, and gene regulatory networks) fall short of universal computation, but are assumed to be capable of explaining cognition and development. I present a class of models that bridge two concepts from distant fields: combinatory logic (or, equivalently, lambda calculus) and RNA molecular biology. A set of basic RNA editing rules can make it possible to compute any computable function with identical algorithmic complexity to that of Turing machines. The models do not assume extraordinarily complex molecular machinery or any processes that radically differ from what we already know to occur in cells. Distinct independent enzymes can mediate each of the rules and RNA molecules solve the problem of parenthesis matching through their secondary structure. In the most plausible of these models all of the editing rules can be implemented with merely cleavage and ligation operations at fixed positions relative to predefined motifs. This demonstrates that universal computation is well within the reach of molecular biology. It is therefore reasonable to assume that life has evolved – or possibly began with – a universal computer that yet remains to be discovered. The variety of seemingly unrelated computational problems across many scales can potentially be solved using the same RNA-based computation system. Experimental validation of this theory may immensely impact our understanding of memory, cognition, development, disease, evolution, and the early stages of life.

+-----+
https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-nbt.2014.0020
IET Nanobiotechnology
Volume 9, Issue 3 p. 122-135
Review Article
Free Access
Self-assembly: a review of scope and applications
Anusha Subramony Iyer, Kolin Paul
First published: 01 June 2015
https://doi.org/10.1049/iet-nbt.2014.0020

Abstract
Self-assembly (SA) is the preferred growth mechanism in the natural world, on scales ranging from the molecular to the macro-scale. It involves the assembling of components, which governed by a set of local interaction rules, lead to the formation of a global minimum energy structure. In this survey, the authors explore the extensive research conducted to exploit SA in three domains; first, as a bottom-up approach to fabricate semiconductor heterostructures and nano-scale devices composed of carbon nanotubes and nanowires; second, for meso-scale assembly to build systems such as three-dimensional electrical networks and microelectromechanical systems by utilising capillary force, external magnetic field and so on as the binding force; and third, as an emerging means to achieve computing via tiling, biomolecular automata and logic gates. DNA, in particular, has been a molecule of choice because of its easy availability, biological importance and high programmability as a result of its highly specific component bases.


Links to my related work

Related links to some of my work are provided below. All of this is in very early-stage development even though some of it has been worked on several times since the late 1990's, early 2000s :

[lists, outline, concept]s to keep in mind from 2020

All of these are very incomplete, but the lists are a handy back-reference so that I don't forget ideas. LibreOffice documents of .odt file format. These are in original form in the directory Mind2020, and while I intend to convert them to html and may update them, I have not done so as of 20Nov2023.