World's longest DNA sequence decoded

[weltweit]

Heard it on the radio, interesting

read it on bing:

A team of UK scientists have claimed the record for decoding the world's longest DNA sequence. The scientists produced a DNA read that is about 10,000 times longer than normal, and twice as large as a previous record holder, from Australia.

read about it in more detail on the BBC:

World's longest DNA sequence decoded

A team of UK scientists have claimed the record for decoding the world's longest DNA sequence.

www.bbc.co.uk

Perhaps in the future prospective parents might have their dna scanned before they decide to have children, I bet there are ethical implications to that. Would anyone pass such an analysis?

Anyhow, it seems sequencing is coming on .. what will it mean?

 

[weltweit]

WHO issues new recommendations on human genome editing for the advancement of public health

Two new companion reports released today by the World Health Organization (WHO) provide the first global recommendations to help establish human genome editing as a tool for public health, with an emphasis on safety, effectiveness and ethics. The forward-looking new reports result from the...

www.who.int

Two new companion reports released today by the World Health Organization (WHO) provide the first global recommendations to help establish human genome editing as a tool for public health, with an emphasis on safety, effectiveness and ethics.

The forward-looking new reports result from the first broad, global consultation looking at somatic, germline and heritable human genome editing. The consultation, which spanned over two years, involved hundreds of participants representing diverse perspectives from around the world, including scientists and researchers, patient groups, faith leaders and indigenous peoples.

“Human genome editing has the potential to advance our ability to treat and cure disease, but the full impact will only be realized if we deploy it for the benefit of all people, instead of fueling more health inequity between and within countries,” said Dr Tedros Adhanom Ghebreyesus, WHO Director-General.
...

 

[Jennaonthebeach]

 

[weltweit]

Was the 16th book I read in 2014

 

[SpookyFrank]

This is engineering rather than science. Being able to sequence even a whole chromosome at once is not actually that useful as the vast majority of that sequence will be irrelevant crap.

I'm not too worried about people's DNA being used to categorise them or do any other sinister shit, mostly because our technical capability is so far ahead of our understanding of how genes actually work and interact. There are relatively few health conditions (or phenotypes that people might wish to select for or against for non-health reasons) that can be definitively traced back to a single genetic locus. And anything involving multiple interacting genetic loci, assuming you can do good enough association studies to definitively nail down exactly which ones they are, might prove resistant to successful gene therapy or gene editing possibly forever.

Humans don't have that many genes, on the grand scheme of things. Probably fewer than 20,000. Our complexity comes from the temporal, spatial and biochemical interactions between the genes themselves, their regulators, their products and everything else those products affect or are affected by. You can scoop up as much data as you like and toss it into the biggest AI brain we've got (this approach has apparently already 'solved' protein folding, but I'll believe that when I see it) but my instinct is that the number of factors will multiply so fast once you get beyond two or three genes that it becomes impossible to solve, like the three-body problem in astrophysics. And just like astrophysics, we might be able to feed in data we already have and create a prediction that plays out the way we see reality play out, but that might be a long way off from getting a predictably different outcome from a different set of conditions. And that is what you'd need to do with gene therapy or gene editing. It's the difference between explaining why Earth doesn't get flung out of the solar system by Jupiter, and building a solar system from scratch and being confident that the planet you choose to live on isn't going to crash into any of the others.

 

[SpookyFrank]

Was the 16th book I read in 2014

I cited it as my favourite book when I applied to universities to study genetics.

I probably thought that was desperately clever of me

 

[weltweit]

Hi SpookyFrank do you think increases in computing power (I forget which law it is, is it Moors law?) will get us closer to comprehending the interactions of our genes? Computing isn't standing still while genes are as they have been for evolution lengths of time.

And there is the potential to grow spare body parts which are fully genetically compatible with the patient because they carry the same gene ..

Will probably come back to this Saturday .. lots of issues ..

 

[Crispy]

Computers can answer very hard questions in a short amount of time so long as you frame the questions in the right way. But we're not even close to being able ask the right questions when it comes to genetics (and gene expression regulation, and epigenetics, and fetal development, and and and).
Just figuring out how basic cells work is mind-bogglingly complex. This is a map of all the metabolic pathways taking place inside every cell of your body, turning chemicals into other chemicals (blue text) using energy from other chemicals (red text) all of it performed and regulated by proteins (blue text) encoded by genes, and all of it horribly contextual and interrelated so it's really hard to study things independently.


source

I don't think it's possible for one human mind to understand it all at the same time. It's my genuine belief that a decompiler/recompiler for genetics will only be possible if we first invent true AI that can surpass us.

 

[SpookyFrank]

Hi SpookyFrank do you think increases in computing power (I forget which law it is, is it Moors law?) will get us closer to comprehending the interactions of our genes?

Closer is probably the wrong word. Computers will help us gain more understanding. That could bring us closer to a workable model we can use to make interventions with specific, predictable consequences; or just give us a clearer understanding of exactly how little we know.

We do make interventions all the time of course, using pharmaceuticals and even gene therapy in a few highly specific cases. But we don't truly understand the mechanisms by which a lot of our medicines actually work, nor why they may work for some people with a given condition but make others worse. But if you give someone the wrong pill and something goes wrong, you can tell them to stop taking it and give them something else. Re-write the wrong gene, and you've made a mistake that can copy itself. With the stakes that much higher, the level of understanding you'd need to make medically and ethically sound interventions at the genetic level is higher.

Results for gene therapy to treat single-gene disorders are promising, miraculous even, but again most disorders with a genetic component are not monogenic. Cancers are never monogenic, because human cells have so many defence mechanisms against becoming cancerous. There are known monogenic risk factors for certain cancers, these are relatively easy to detect and potentially easy to fix with gene therapy. But if your only aim is to remove a risk factor rather than cure a known disease, and your method for doing so comes with its own set of risks as any gene editing always will, the maths starts to look wobbly.

By the time you get to conditions in which three or four genes are involved, I think the level of complexity is not just prohibitive now but always will be. Say you find four genes that associate with a disease, A B C and D. Let's call the 'bad' alleles (alleles are different versions of a particular gene) A' B' C' and D'. Well it could be that the D' allele is actually protective against the condition and people without it don't survive the first week after fertilisation if they also have A' B' and C', and that's why the D' allele always seems to show up in disease patients. We might use gene therapy to 'fix' all four alleles, and actually make matters worse. And you can't do clinical trials in the same way as you do with pharmaceuticals; by testing them first on healthy people. It wouldn't be ethical to do so and you wouldn't get any useful data anyway.

I defer to Crispy on the computer stuff because that's not my field. But I'd agree with him that we're not even close to the point of being able to feed all the relevant data into a computer, never mind get useful results out. When you're talking about individual molecules affected by quantum-scale processes, we may never be able to get sufficiently fine-grained information about what's going on in a cell to create accurate models even with limitless computing power. You eventually hit a brick wall in terms of what the universe will allow you to measure. The more interactions are involved, the greater the effect tiny differences in conditions can have on eventual outcomes. I reckon we'll have a convincing computer model of a working human brain before we have a reliable computer model of the genetics and biochemistry of a single living cell.

 

[weltweit]

I was thinking that this sequencing of a long dna gives hope for further development of things like RNA vaccines which seek to align with a genetic target ..

However I defer to you guys because my knowledge is very basic ..

However

Pfizer coronavirus vaccine: What is RNA and how does it work?

The Covid-19 vaccine remains a huge talking point following the Pfizer and BioNtech breakthrough - so what does RNA mean and how does the vaccine work?

www.lbc.co.uk

What are mRNA vaccines and how do they work? – UK Health Security Agency

The official blog of the UK Health Security Agency, providing expert insight on the organisation's work and all aspects of health security

ukhsa.blog.gov.uk