🧠
brainbrane* ("
lnq") is our
kernelspace@ (ie.
scalable^ technopathic n-manifold) 🧠brain [
= central nervous system (cns) / nerve net (nn)]
reagent to
regulate+control cell/
neural signals, among other
userspace functions.
** Fibor(s) as a cell matrix. @@ Printed from. ^^ Mapping a neural network.
/// +🧠'brainbrane (bb)' would be the generic name for the 'lnq' technology (reagent = natural logarithm (ln) of the reaction quotient (Q) of the species involved), differentiating it from the person.
+We should think of the reagent as a propagating kernel (the 'genetic code' that directs information+translation in our userspace [🧠brain] example). Here, 'kernelspace' (= proof of the Quantumquotient) is any gate that can be debugged and/or have its processes interrupted, and 'userspace' [host] is the address registrar (ie. what registers some subset of some kernelspace). The two form an aggregate operating system.
+In this case, it is a drug💊 (ie. organic material determining physiological responses by regulating any and all internal sinusoidals) whose biological target is the entire 🧠brain.😵🤯
+This is colloquially known as '🧑🏿lnq's drug', and is also offered as a class of tangible proofs.
As the formal outlier of
gameplay mechanics (ie. think "
gameboard" - standard board layout for
juking), this is
UUe's
interface/
jukebox kernel augmenter. This
neurally networks (ie.
maps@)
🧑🏿lnq's walk.
Jukebox instancing* utilizes
portable,
peer-to-peer core batching (ie.
cellular nodes on
dial)
~ for
walk-
tracing and
execution.
@@ A type of hyperlinking ** linker ~~ neuron
The
range of the
threading (being essentially the reach of
orchestration^, and, hence, the network) is
determined organically.
^^ 'orchestrate' means sequence and map. Cybernetically, this is a set of
automata augmenting (ie.
stew patchwork)
gameplay. (see
joey,
Egglepple, everywhere!,
BOT)
/// Network growth means that scalability in spacetime complexity is sought. In jukespace, this may be virtual, physical, or hybrid.
+Hypothesis: kernelspace and userspace are self-similar.
open framework
outline
Steps: (
1)
preimage the protein, (
2a) construct a
distributed supercomputer [lnq] to (
2b)
juke the polymer, and (
3)
solve the structure with machine learning (
genetic algorithms).
Theorizing that
DNA/RNA/protein bindings should be
algebraically solvable,
we are
juking certain classes of macromolecules for better
models+investigation of their related neuropathologies (sampleset: PD, HD, MS, Alzheimer's/diabetes, ALS), cotangent viral groups (sampleset:
EBV, HIV, influenza, Ebola), and possible extrapolation to cancer🎗 (ie. kinase catalysis). Doing so requires decomposition of a large number of long
random walks, to which we are applying a homebrew
artificial intelligence (
UUe.i) that
runs a huge
machine learning algorithm. Our
compute is distributed over a mobile
cluster, allowing for
user engagement in
PP. Furthermore,
cinematic properties are addended for driver monitoring; assuming that the eye is an extension of the brain, we gather (
using the camera system on mobile devices) large
batches of photographs (still,motion video) of the human eye(s) in various settings/strains for
mapping α-syn and
β-amyloid (Aβ) plaques (+ other proteins).
My
solution (
theory,
proof-of-concept) for
imaging the protein is to treat intracellular
folding pathways (trace) like an animation exercise, where plaque in the human retina is photographed at
start/stop endpoints with an intermediate
random walk acting as molecular
frame interpolation. Conditioning the polymer to behave like a self-driving vehicle, we
use fleet learning (crowdsourcing) for molecular/cellular navigation. A
compression algorithm is then applied. This relies heavy on
artificial intelligence [
ai] (ie.
genetic algorithms for predictive analytics applied to biopolymers+
biometrics) and involves utilizing MMO/
crowd computing for purposes of
gamification, versus traditional labwares. Development is
open to the community as a way to fast-track
builds, and to honor its promise of
UUe being a public utility
en perpetuity.
We are not creating a chemical (tangible pharmaceutical) nor a therapy. This is strictly molecular modeling of peptide conjugates using distributed machine learning, the preliminary step in accommodating lab-based rational drug design+synthesis.
We conjecture that the above technique is applicable to any and all macromolecular determinants.
abstract
by
🧑🏿Link Starbureiy et al.
The aggregation mechanism of
alpha-synuclein (aka
SNCA/NACP gene
encoding) is
undetermined, as to be expected since the protein is intrinsically disordered (ie. chemically unstable). It has been suggested that this unstructured tetramer is a combination of
alpha helix and
beta sheet conformers in equilibrium. There exists lab-produced evidence that the protein interacts with both
phospholipids and
tubulin[j] within
presynaptic terminals to form (characteristic) Lewy bodies, thus, it is a pathological component in
multiple system atrophy (an umbrella term
used here for would-be neurodegenerative [inverted dopamine modulation] diseases).
We wish to have a
walkable model of synuclein. Current popular
molecular dynamics programs on the market
[a][b][c][d] are strictly brute-forcing
chemical force fields, and
not accounting for
twistorspaces, thus missing a more potent subset of polynomials in the
protein folding problem. Because medical resonance techniques (especially those in radiolabeling) require
datasets of
preimages/
images, it only makes sense that a positron emission tomography (PET) tracer (as requested) should be constructed from a camera utilizing
volumetric* pixel-level signal compression. For this (the signal sampling scheme), we'll
use the
Stewdio to
hash EGP. Right now, the
api is public and the
math works, but it needs substantial
computing power for realtime implementation.
with depthnets+segnets
The secret to curing any disease is with
cell division (ie. iteration). Think of it in the way of an automotive analogy -- if you wanted to teach a car how to drive itself, the instinct may be to
program its 'eyes' (camera+radar) to adhere to
maps. A smarter way is to just let the car follow patterns it has learned from other drivers. In other words, the best way to train the system to drive is to simply show it what driving actually looks like by creating an agent that
clones behavior. That's how
computers are trained; give them
big data sets of what is correct and incorrect, and then
write a
classifier. Which is
precisely what we are going to do here (in our case, include
animation modifiers); take live
strategies (an
array of preimages, ie. divided quanta) derived from
juker activity [
genetic algorithms] and apply them to known protein structures (specifically,
SNCA,
parkin,
HTT, and
LRRK2). Over time, we'll get
tighter bounds on
errors that will eventually only produce optimized images.
This is done
using UUe, a 'smart' (
artificially intelligent)
preimage-image compositor native to
distributed computing (ie.
distributed memory). The
ai will do two (
2) things here: (
1) be a background
embed that
pulls+loads idle
compute resources for
processing, and (
2) provide visual ventriloquy [
+training] for image
classification+mapping onto an actor [ie.
string] in the
capture pipeline. It is designed as a
distributed framework that
uses the
processors from mobile devices (eg. phones, tablets, VR goggles, watches, etc.) to power the
compute. Today's machines tend to be more
game-centric, so the
GPU is a premium, which is good since the
ai will be performing specialized
3D renderings. It can augment
opensource libraries (eg.
OpenMM,
+GROMACS,
BOINC,
TensorFlow) in addition to the
Stewdio. In
theory, this will be
the largest and most powerful supercomputer ever constructed. Once ready, it will be freely available for
uses similar to this
one en perpetuity.
significance
No doubt, drug companies have done astonishing work thusfar considering that their modalities for quantitative predictability of synthetic chemicals and bioorthogonals have very limited
data-driven
processes for
determining structural motifs (assumed macromolecular architecture). The proliferation of genome sequencing projects (eg.
Human Genome Project) simplifies cellular polymer/protein primary structure (amino acid sequence) inference. However, that information doesn't reveal much about protein
function, which is necessary for our ultimate goal of rational drug design (albeit via
computational biochemistry). That comes from knowing macromolecular
three-dimensional tertiary structure. It is also important to note that protein folding happens in solution versus isolation, so any serious contender for a
structural motif must incorporate at least hydrogen (usually with oxygen: H + O
2) in its formula (ie.
amino acid residues are surrounded by water). This is because protein structure conforms to thermodynamics (/energy landscape)
[w]; as temperatures drop, said structures collapse, whereas in higher temperatures, hydrogen bonds [amino acids-H-
residues] break and re-form (mimicry of
symmetry-breaking).
UUe.i can dramatically reduce discovery-to-classification time (and thereby remove the evaluation bottleneck) precisely because it is (in
theory) the heaviest duty
program on the market geared towards molecular dynamics and
numerical analysis. To satisfy a certain rewards committee, this is being advertised as a radiotracer
~, but on a deeper level, it is a generic
solution to
protein structure prediction, meaning that it has much broader application (biology, ecology, etc.) for consideration.
~~ A radiotracer is a chemical compound which, having dropped (proton) or absorbed (neutron) an isotope (radionuclide) from an atom, behaves as a mechanism for nuclear reactions/chemical processes in accordance with radioactive decay. In laymen terms, we're highlighting only part of a chemical element (say hydrogen) to track it through various pathways, at the same time revealing different imagery. The protein is traced by voxelating its longitudinal pathways in the retina using the camera system on a mobile device. Another bennie is that the
program is at the forefront of scientific innovation in regards to
supercomputing.
Solutions to
protein folding and similar RNA folding problems are largely
computational. Engineering+maintaining a
supercomputer is expensive, but the payoff is near-immediate and widely
useful for everyone. This is probably going to be a long-term endeavor, which is
why the project has been opened up to the community at-large. Implementation needs boil down to underwriting.
🧀Back this to enable development (improved core features,
bug fixes) continuity.
Function map:
🧠brainbrane ⇔
avatar
