I mean, in theory that's possible.

But in practice: personalized medicine is impossible to tests. We rely upon giving tens of thousands of volunteers medicine ahead-of-time to prove if medicine is safe.

While the technology theoretically exists to make and distribute personalized medicine, the ethics and safety questions of doing so remain unanswered.

Personalized medicine has been tested for years in clinical trials. Trials require tens -- not tens-of-thousands -- of volunteers.

Personalized therapies are being used to treat cancer patients right now, and using machine learning to find the right binding site for a particular patient's tumor should present no more ethical questions than giving them NSAIDs.

Uhm yeah, but it also opens up ways for hackers to code all sorts of biological stuff and put it in our bodies. All from behind their keyboards.

Imagine "if dna.get_race() == $RACE then kill();"

Or how about some biological ransomware?

That's really not how mRNA works. If DNA is the source code of our body, then mRNA is the machine code that exists in L1-instruction cache. (There's another process out there that copies DNA into more readily processable RNA)

The way any of this mRNA stuff works is by throwing instructions into our body to create certain proteins in certain configurations. For COVID19, the vaccine is... as XKCD-put it... a set of blueprints to build a "fake death star" without any weapons activated.

https://xkcd.com/2425/

------------

mRNA further has a "innate safety" mechanism, in that it degenerates. That's why our body uses DNA after all: because DNA does not degenerate, even though RNA is what's actually executing so to speak.

So any mRNA medicine will have to be strictly temporary, and get its job done in a limited timeframe. The COVID19 virus gets around this fact by self-replicating. The instructions that our cells execute are to create a new COVID19 virus. The original "quine", COVID19 (and all viruses) "print themselves" as part of their execution.

-------------

The "vaccine" is a set of blueprints for the COVID19 "spike protein" (and ONLY the spike-protein). Since it is missing all the other parts of COVID19, it cannot self-replicate. Our body then gets trained on recognizing the COVID19 spike protein, and is ready when the real thing attacks our body.

I believe there are also viruses who can insert themselves into the cells DNA, and there are some articles on how some parts of human DNA is already old virus genome. Viruses are far more dangerous in that regard.

There's both DNA viruses and RNA viruses. RNA viruses are way more common though.

> [DNA] Viruses are far more dangerous in that regard.

Not necessarily. "Corrupt DNA" can be somewhat detected the body as a cancer cell. Our "Natural Killer" cells then kill those cells.

Cancer / corrupted DNA happens all the time in our bodies. Even healthy bodies (!!). The difference between a cancer-patient and us however, is that a cancer-patient is overrun with cancer-cells.

Our bodies naturally kill off cancer under normal circumstances. Figuring out why cancer / corrupted DNA completely takes over the body is a big mystery, especially because our body is so good at fighting off cancer under normal conditions.

> then mRNA is the machine code that exists in L1-instruction cache

But we all know that once an attacker has write-access to the machine code in the cache, all is lost ...

I mean, my point is that "if (race == whatever)" is completely nonsensical from an mRNA perspective.

There's no way for mRNA to scan the rest of your DNA sequence to make an if/else determination. mRNA is just gonna execute once its in the body.

> Or how about some biological ransomware?

Yeah, that's called a poison and antidote / antitoxin. You don't need mRNA for that. Poison someone's food, and as they lie dying, you can offer them the antidote in exchange for something.

Also, one would have to define "race" in genomic terms. Good luck to the person attempting to do that.

Is that really that hard? A $100 genetic test will tell you what race(s) you are.

Will it? Sure, you can predict from where your ancestors came, haplotype, etc.

But define race in a genomic sense. It's a feat that key genomic experts have not yet been able to accomplish. For instance: http://ewanbirney.com/2019/10/race-genetics-and-pseudoscienc...

> I mean, my point is that "if (race == whatever)" is completely nonsensical from an mRNA perspective.

Is it? mRNA can code for a protein that detects other mRNA in the same cell, which may reflect any genetic characteristic of the host.

mRNA isn't really code. Its a blueprint.

https://www.nature.com/scitable/topicpage/translation-dna-to...

When we say mRNA "executes", that's a cell injesting mRNA, and assembling a protein (polypeptide). For example, the mRNA sequence 'ACU', when 'executed' by a cell, will turn into a Threonine: https://en.wikipedia.org/wiki/Threonine

RNA can fold up and get stuck to other sequences. Using that as a trigger isn't out of the question.

> There's no way for mRNA to scan the rest of your DNA sequence to make an if/else determination. mRNA is just gonna execute once its in the body.

Of course there is, genes can be turned on/off using transcription factors. Still, the sequence of vaccines are published and easy enough to sequence and analyze, so I don't think something like this could be mass produced.

DNA = X86 opcodes RNA = UOPS/ROPS

A better analogy would be anonymous functions thrown into the global namespace that get called with whatever arguments that may happen to be on the stack at that moment.