Hacker News Reader

Not sure how much ASLR would be relevant w.r.t. genomes and cancer...

Malicious hackers are on the one side very smart (the cancer genesis process is not smart at all), but on the other side they make poor use of evolutionary strategies, mainly because human hackers can't easily spawn "an army of clones of their mind" (cancer otoh is all about evolutionary strategy).

Since the information about "what mutating a codon in a gene does" is not really available to "the attacker", it's as if the whole "address space" is already randomized by default...

The "cancer problems" sound like it's more like being attacked by a really really really dumb hacker, who has access to an utterly insane amount (relatively speaking) of computing power!

(...kind of hard to defend against such an attacker unless you invent modern crypto stuff like hash-signatures, blockchains etc. And our "DNA computer" is not at all designed for these types of math, recoding it to do this would probably be harder than developing mind upload capabilities or superhuman-AI :P Hence not likely to happen.)

At one point during the cancer progress process the cancer is characterized by genomic instability (high mutation rate) and extremely high metabolism(high replication rate). The immune system is actually attacking the cancer cells successfully(in generalizing here in regards to cell types), and this may lead to an optimization process where the immune system is selecting for a cancer cell able to avoid the immune defenses. This interplay creates a microenvironment which is as you described, evolutionary in nature.

As for cryptography, I would argue the biology contains hash signatures and other configurations common to cybersecurity(blockchain turns one problem into many IMO). While there may not be asymmetric cryptography, there is also no pseudorandom business.

I have no idea what you are talking about but does human biology have computation at the core? Any papers on these?

biology have computation at the core?

Doubtful. You can to computing with DNA http://dna.caltech.edu but it is still debated if the actual role of DNA is being a code (some recently weasel out form that stance by saying it's an “app”, like there's a difference, machine code being a code). Coding theory applied to DNA yields inconclusive results. Galois theory usually has power over any kind of information encoding, cryptographic, computable or not. One constructed one for DNA convinced mathematicians it's not the way to go at all. If it's computation it's nothing like what we mean by any model computation, you may as well say it's magic instead of making stretched analogies.

About OpenBSD and evolution, that's so fetch... these folk invented the attack in the first place, not evolved a response to some market force in the early oughties. Broadly I don't think designed systems are in business of evolving, otherwise living organisms could perhaps evolved electro-hydrostatic instead of hydraulic power system with a pump being a single point of failure.

Linking biology to specific concepts in computer architecture or worse, with OS design, is indeed a stretched analogy but biology is rich with computation. I'll try to be short.

A cell must sense and respond to its environment. One aspect of this is regulation of gene activity by interactions of for example, transcription factors. When interaction types can be well enough approximated as either of inhibition or activation, you can model them with boolean networks and when levels matter, some have found moderate success with recurrent neural networks under a restriction on what NN nodes represent, to ease interpretability.

That is not the same thing as saying your genome unrolls neural networks. What it is actually saying is that the complexity of the best performing neural network indicates the richness of the underlying computation, and the predictive accuracy of the model captures the functional equivalence of the respective computations (in the cell and in the neural network). The learned model is the instance and neural networks are the class, it is a mistake to place emphasis on neural network, they are merely a way of packaging the computational model of interest.

A cursory review here: https://www.sciencedirect.com/science/article/pii/S001048251...

If you have time for videos: https://www.youtube.com/watch?v=vA9D727AGHI , https://www.youtube.com/watch?v=ZX-GrO2qXQM

More relevant to this article is the study of cancer within the framework of evolutionary game theory, as done here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2768082/

There is a precise correspondence between that formulation and one done in terms of online sequential prediction or the so called experts algorithm. All of these are models and still quite limited at that but biology is complex and the future lies in the development and leveraging of these rich connections.

Can you explain how the analogy is stretched? It’s my personal belief that biologists tend to be fairly implementation focused from how they learn about the systems involved.

In biology there is 2 factor authentication to protect against parasitic spoofing, intrusion detection systems, protected kernel memory, memory hierarchy, policy based permission. It a very nice operating system IMO.

Analogies can help but if taken too far, will provide only illusionary knowledge from superficial similarities. Consider, if you trace out the hierarchies induced by interactions in biological networks and then do the same for an OS's call graph, you find quite different topologies reflecting their different priorities. In computers, in part, efficiency and reuse. In biology, robustness, minimized interdependencies, developmental stability and more.

There are things that concern our electronics that do not matter for biology and there are many things biology must allow for that our hardware cannot. Tying things down to the peculiarities of our systems too narrows the scope of applicable models and understanding. Sometimes it is more useful to think in terms of stochastic differential equations than in terms of operating systems.

DNA compilation is pretty Read Eval Print Loopy to me...

dna replication is very map-ish. then they can fold back to a reduced state

Well all info can be encoded as bits (Shannon’s theory of information) and the genome isn’t very far from that encoding. You can divide with bitshifts with computers, it’s possible it could be occurring in biology(this area has been looked over in analysis until recently).

The properties of the genome are similar to the storage memory in a computer in that malicious code is inserted into the path of the biological interpreter. Early during human embryonic development (and only during this time for the human) you have a sudden increase of retrotransposon activity. The DNA itself is like stored polymorphic computer virus with a compiler and code to interface with the environment. It’s very unusual the timing and genes carried around by this circular process are vital for life. That is my basis for the comparison to ASLR. There also exists 2 factor authentication, signature based intrusion detection, intrusion prevention systems.

Here is an interesting paper on an organism that has high rates of this DNA recombination..read the motivation section and you decide if this is computation.

https://ac.els-cdn.com/S0166218X09002534/1-s2.0-S0166218X090...

IDK about papers per se, but as a computer scientist who took classes in bioinformatics, there are parallels. At the base level, cells are machines for constructing proteins, and information on that process is encoded in DNA. Silicon uses 2 bits, DNA has 4 base pairs. Those base pairs groups in sets of 3 called a codon, kinda similar to a byte. Word size is normally a multiple of 8 for computers, but the 6 bit word in DNA is more than sufficient to cover the 20 amino acids, leaving room for both a stop sequence, and some rudimentary error correction -- a mistake in copying DNA somewhere might not end up changing the resulting protein.

There's also a level of programmatic structure in terms of transcription promotors and gotos, and probably some level of conditionals. DNA is also special in that it can be read forwards or backwards, and the base pairing means you always have two potentially transcriptable sequences, each starting from one of 3 offsets. Imagine if busybox implemented all of coreutils in one binary by shifting the word alignment 1 bit to produce a completely different program.

Finally, at the intracellular level, we have a whole host of message passing between cells using hormones and other signaling molecules, and organs like the pancreas that try to regulate a system using them.

As far as 'computation' I don't think the system is Turing complete, and the systems aren't designed to answer decision problems, just construct things, possibly in response to an environmental factor.

> (I don’t think this compressed state can be examined, but compression algorithms are seriously under appreciated)

I once attended a lecture at a physics conference by a guy who basically used gzip to do sequence alignment (using compressibility as a way to measure Kolmogorov complexity). Might be interesting to you. The title was: Blasting and Zipping: Sequence Alignment and Mutual Information

I have no idea what the heck you are talking about but is human biology have computation at the core?