I think it's certainly possible to get mixed or even bad results with gene editing by trying to push the technology farther than we really understand.
That being said, the basic priority for gene editing is probably to prevent/reverse bad genetic conditions that we understand well. No more downs syndrome.
That seems to be an unmitigated win to me.
DNA looks mightily like an extreme and profoundly complicated, complex piece of programming, with unnecessary redundancies removed, and useful redundancies left in, as well pieces of code that call lots of other pieces, as well as being called by others.
If the genome is analogous to a computer binary... we don't understand the file format, and how it is read and interpreted.
Nature knows far-and-away better than we do how to make it all work 99.99% of the time, with occasional errors due to radiation, poisoning, interbreeding being accounted for so that things don't get out of hand.
Well, these assumptions held... until we intelligently stupid creatures came along, using our intelligence to create dangerous tools that are most abused by those least worthy to wield them...
Nature does not ‘know how’ to make it work 99.99% of the time. Nature very frequently aborts pregnancy’s when any number of things go wrong. Getting all the way to birth is a huge checksum that validates the majority of critical DNA.
Which gets it the larger point. Evolution focuses more on species propagation than individuals.
Cell division is one of those things you need or get right or you don’t get a pregnancy.
It’s not like the mother needs or do anything at that point, the result is just death by default.
We may understand a lot, but nowhere near enough to be able to confidently edit DNA, and know that it won't causes a ton of unforeseen issues down the line.
We don't know what we don't know ~ that is, as much as we like to believe that we understand a lot about DNA and how it works, odd things pop up all the time that prove us wrong. I dare say that there are far deeper mysteries about DNA that we have no current understanding of.
> Nature does not ‘know how’ to make it work 99.99% of the time. Nature very frequently aborts pregnancy’s when any number of things go wrong. Getting all the way to birth is a huge checksum that validates the majority of critical DNA.
This is merely an example of nature getting it right ~ abort if there are serious issues.
> Which gets it the larger point. Evolution focuses more on species propagation than individuals.
It is about both ~ the individual matters most ~ species propagation doesn't even register in importance to a majority of individuals. Having children does. They're subtly different ideas.
That’s a question of code not file format. I may understand C, but that does not mean I can edit any C program safely.
> abort if there are serious issues.
That stance suggests there is no way to fail as death is the worst option. But even that fails as a significant number of women died due to pregnancy or birth before modern medicine. Even now pregnancy’s are higher risk than most things people do.
People with personalities also have neurons and wonder of wonders a centural nervous system.
PS: It’s hard to get good numbers for this stuff as but many things like drug use my pregnant mothers can cause significant developmental issues. But, everyone has several random mutations and mental problems from them seem to be far less common than other issues. Especially when you exclude things getting passed down over time and just look at new mutations.
I would add that PheWAS is often performed with specific genetic instruments of interest, and can be performed with aggregate measures such as a genetic risk score. This lets you see association across a slew of traits at once.
And it's good we are sharing this info. But we are far from having high-powered GWAS for all of human's traits and diseases. And often what we do know is from a single demographic (fwiw mostly caucasian, not chinese). How gene variants influence outcomes in different genetic backgrounds, or within certain environments (think developmental nutrition in africa vs asia) is typically a big questionmark. So IMO performing CRISPR edits to mitigate a problem that was completely avoidable by other means was unethical at best.
For UK Biobank, there is a GWAS of ~everything: http://www.nealelab.is/uk-biobank/
The lack of diversity in genetic studies is a huge problem. This one isn't just moral, either: we're ignoring the richness of allelic diversity that is found in people throughout the world (particularly in Africa, which has the most genetic diversity because of its pre-bottleneck status). There is an awareness and I think you'll see big changes here in the next few years.
Many people are interested in seeing CRISPR push forward, and mention something like you have: but what about the "knowledge it will provide us for the next iteration". Trust me, I'm just as eager as anyone. However, if this mavrick experiment fails, and something is remotely off about these little girls (u can bet they will be under a constant magnifying glass), this will end up being a huge setback. It will spook the public, and the legislation factory will churn out regulatory policy that nobody wants.
It's dangerous to tinker without a truly full knowledge of what one is doing.
CRISPR Twins Aborted