Thinking about big science, the human genome project looms large. I'm not an expert so: did this project fulfil its potential? From what I've seen, it seems the major achievement was development of sequencing tech and a focus on open data. Otherwise, people still talking in terms of potential value?
Apologies in advance if this comes across as ignorant, I really don't know much about genetics.
=> More informations about this toot | More toots from neuralreckoning@neuromatch.social
tangential anecdote:
During the project, lots of bloviating about how only one approach would be validated by The Win. But in the teens I was in a group including midlevel scientists from both “sides”. And they all said that each approach depended on something the other was specializing in, and they had been sharing intermediate results.
@neuralreckoning
=> More informations about this toot | More toots from clew@ecoevo.social
@neuralreckoning I don't have any sort of complete answer for you but I think you would enjoy reading this if you haven't yet: https://www.genome.gov/virtual-exhibits/human-genome-project-is-simply-a-bad-idea
=> More informations about this toot | More toots from wronglang@bayes.club
@neuralreckoning I'll add that it had some of the same important effects every other big science project has had: the consistent training of many researchers at all levels in a field that society has decided is important to fund. Was definitely a better choice than the current AI rampage.
=> More informations about this toot | More toots from wronglang@bayes.club
@wronglang thanks for the link! I tend to agree about the level of spend on AI but I guess if it works and if the corporations don't turn it against us, it might turn out to have been worth it?
=> More informations about this toot | More toots from neuralreckoning@neuromatch.social
@neuralreckoning one positive I've seen is the whisper model: for English it's a turn-key model packaged for all sorts of uses that does speech to text with surprising reliability (for at least some voices) so like for free software assistive technology it seems like a real blessing. All the previous models I had tried were just monstrous to get going with.
=> More informations about this toot | More toots from wronglang@bayes.club
@neuralreckoning I'm also no expert, but I did just stay at a holiday inn express see a lecture that discussed this a few weeks ago.
The HGP was crucial to advancing the science of human genetics. One of the most important things we learned via the HGP is that genetics are way more complex than we thought. A lot of things that were believed to be "a gene" are mutlifactorial, epigenetic or something else. It led to fewer breakthroughs than expected, but did create important foundations.
=> More informations about this toot | More toots from plaidtron3000@jorts.horse
@neuralreckoning yes hello, I do know a bit. The human genome project can now be reproduced in a single week. We now have a depth and breadth of DNA knowledge of all kinds of human beings the original project never dreamed of. But to get to where we are today the HGP had to happen first. One of the main discoveries may have been that our genome is far more nuanced and complicated than people thought. There is no chance we would not have done the project. It was inevitable. 1/
=> More informations about this toot | More toots from bert_hubert@fosstodon.org
@neuralreckoning We might have hoped for more immediate breakthroughs, and people were surely expecting those. These days we are working on 'pangenomes' which incorporate all human DNA variation, and these may in fact hold more 'smoking gun' results than the "universal genome" created by the HGP. But because we would not have gotten where we are today without first doing "a" HGP, it has by definition been a success. /end
=> More informations about this toot | More toots from bert_hubert@fosstodon.org
@bert_hubert right but that's the technology development I mentioned in the OP, right? We can only do pangenomes because of rapid sequencing tech?
=> More informations about this toot | More toots from neuralreckoning@neuromatch.social
@neuralreckoning @bert_hubert
I think the previous poster's comment that we could do the whole project in a week now is misleading. Yes, we can sequence an individual genome that quickly now, but a lot of that is because we can compare the newly acquired data to the template provided by the original HGP. Everything we do with human genetics now builds on that project.
=> More informations about this toot | More toots from VATVSLPR@c.im
@VATVSLPR @neuralreckoning very good point, I should have taken that into account. The HGP was high pressure lots of money and prestige. With less money we’d also gotten there but maybe a decade later. And that would have set progress back a lot. So in that sense too it was successful.
=> More informations about this toot | More toots from bert_hubert@fosstodon.org
@bert_hubert @neuralreckoning
I think the immediate benefits of the HGP were oversold. People talked about it as the key that would unlock our understanding of human biology, but it was like the key that unlocks your front door. You have to get out of the house to explore the world, but the outside world is unimaginably bigger than you expect when you've spent your whole life living in one house.
=> More informations about this toot | More toots from VATVSLPR@c.im
@bert_hubert @neuralreckoning
“more research is needed” (not a bad thing, but not the original goal, see also blue brain, chan zuckerberg, etc)
=> More informations about this toot | More toots from knutson_brain@sfba.social
@bert_hubert thanks for the reply!
"One of the main discoveries may have been that our genome is far more nuanced and complicated than people thought."
But that sounds like a fairly high level point of view that we probably would have got to without doing it? Maybe that's an unfair response given that we don't know that we would have found that out without HGP but we do know that we got to it by doing the HGP.
"There is no chance we would not have done the project. It was inevitable."
Right, in the history I read this was one of the points made early on in discussions: it will happen, it's just that it will take 30+ years if we don't prioritise it. Does that make it a good thing though, necessarily? I don't know.
=> More informations about this toot | More toots from neuralreckoning@neuromatch.social
@neuralreckoning @bert_hubert
This is less true than you'd think. The HGP was heavily criticized for sequencing all the "junk" DNA biologists didn't understand the function of. It turns out that "junk" has important functions we didn't know or even dream about. We might eventually have understood that stuff without the HGP, but it would have taken a lot longer. Sometimes you really learn more from one big, groundbreaking project than from a thousand little projects.
=> More informations about this toot | More toots from VATVSLPR@c.im
@neuralreckoning
As the HGP was being completed, we saw the same sort of personalities and competition and certain invective that we see with all big-money projects. We saw the same sort of promises being used to secure the money, and then those promises being forwarded to the public (by governments) or shareholders (by the privates) to justify the expense.
So that is interesting.
=> More informations about this toot | More toots from red_concrete@mastodon.social
@neuralreckoning
The project itself was a very humdrum piece of science, very boring. Hence much of the hype was about speeding it up, who would finish first. I never quite understood how you could then justify speed of sequencing as one of the pluses to emerge from it.
Maybe it is a plus in the long run, but pending actual benefit from knowing the genome sequence, that is sort of putting the cartbefore the horses
=> More informations about this toot | More toots from red_concrete@mastodon.social
@neuralreckoning
In terms of the benefit of the data, and all the hype around the medical advances, not a lot has planned out IMHO.
Rare and monogenic diseases have certainly been mapped, but you can often enough do that without full sequence.
Common and chronic disease (where you can get real bang for your buck in terms of population health) has turned out to largely not be dependent on one or two or even a few genes. Largely as expected, really. Clinical benefit has been slow for all..
=> More informations about this toot | More toots from red_concrete@mastodon.social
@neuralreckoning
What's striking is that the volume of data, the development of molecular genetic tools, of computational tools has put innovative, mechanistic work in the shadow. There hasn't really been any "great leap forward" in terms of new insights, new laws of life, and so (am I right on this, happy to be corrected) no Nobel prizes for such stuff, that could be attributed to the HGP.
=> More informations about this toot | More toots from red_concrete@mastodon.social
@neuralreckoning
There's been utility of the sequence in archeological genetics, comparative genetics (across species, insights into evolution), and in the (slightly more interesting) slog of annotating the sequence (where are the genes, how are they different, what types are there, how do they seem to work).
Some of this matches the pre-2000 hype and promises, but not on the most impactful fronts.
=> More informations about this toot | More toots from red_concrete@mastodon.social
@neuralreckoning
Of course a negative result in science is still a result. It's interesting that diabetes is so complicated genetically, and yet turns up as quite heritable in family studies. Psychiatric disease even more so. There might be general high level insights there, but no one has grasped it, or operationalised it.
Also. Some areas of human medical genetics are out of my field, so e.g. cancer genetics has advanced faster, and maybe I don't have a sense of that....
=> More informations about this toot | More toots from red_concrete@mastodon.social
@neuralreckoning
Cancers and some other diseases had relatively well-charecterised molecular bases, independently of the HGP, and that was boosted. But who is to say by how much -would advances in those diseases have been made anyway, without whole-sequence data or tools?
=> More informations about this toot | More toots from red_concrete@mastodon.social
@neuralreckoning
I would support your hunch on the outputs from the genome project.
You get plenty of people who are believers, though! I've been shouted down (on 'other platforms') on occasion, I admit.
But the promises now touted are still in the future, usually hooked onto new analytic methods etc which have yet to materialise. Always the next best thing....
That said, I am a believer! I think it has all that potential for human health, and more! Just no clear path towards that, yet.
=> More informations about this toot | More toots from red_concrete@mastodon.social
@red_concrete thanks for that! It's really interesting to see different scientists' impressions of the impact. Very eye opening to me, a non-specialist.
=> More informations about this toot | More toots from neuralreckoning@neuromatch.social
@neuralreckoning
Without the Human Genome Project, we wouldn't have a comprehensive set of protein sequences for our species. My career in clinical #proteomics would have faltered at the starting gate without being able to match tandem mass spectra to corresponding peptide sequences.
=> More informations about this toot | More toots from dtabb73@mastodon.africa
@neuralreckoning
We use the results of the project All The Time. As other commentors have already pointed out, we use subsequent generations of data more, but some of the techniques used to get those subsequent generations were themselves dependent on the output of the HGP. (And not just the tech dev, but the data; we assemble a new genome sequence now using the existing "reference" sequence as a starting point.)
( … )
=> More informations about this toot | More toots from stevegis_ssg@mas.to
@neuralreckoning
I struggle to convey to my younger colleagues what pain was involved in figuring something out that you can now just look up. "Chromosome walking" was a thing and I've done it and you do not want to. Many modern techniques also have the existence of the whole genome sequence implicit in them; we can map ChIP reads and call peaks, we can select crispr guides to minimize off-target effects, only because we know a whole genome sequence.
=> More informations about this toot | More toots from stevegis_ssg@mas.to
Was #Inria involved in HGP where the main implementation of #Guix for reproducible research in bioinformatics was urged?
@khinsen @civodul
=> More informations about this toot | More toots from sharlatan@mastodon.social
@sharlatan To me knowledge, HGP is older than Guix and it’s more CNRS than Inria; I don’t know if INRIA had joined the HGP party.
@civodul @khinsen
=> More informations about this toot | More toots from zimoun@sciences.re
@zimoun HGP ended in 2003, Guix started in 2012. In France, CNRS and CEA participated in HGP. I don't know about Inria.
@sharlatan @civodul
=> More informations about this toot | More toots from khinsen@scholar.social
@khinsen @sharlatan @zimoun What’s HGP? :-)
=> More informations about this toot | More toots from civodul@toot.aquilenet.fr
@civodul Human Genome Project. More info here: https://en.wikipedia.org/wiki/Human_Genome_Project
@khinsen @sharlatan @zimoun
=> More informations about this toot | More toots from Guillawme@fediscience.org
@Guillawme @zimoun @sharlatan @khinsen Oh thanks. So a deployment tool was developed in the context of HGP?
=> More informations about this toot | More toots from civodul@toot.aquilenet.fr
@civodul A deployment tool specifically, not that I know.
But a lot of bioinformatics things were developed for this project. A prominent example is genome assemblers, both the implementations and the underlying theory. They are needed because of the way the sequencing is performed: it's called "shotgun sequencing" and produces many short overlapping fragments (called "reads") that are then assembled into the final sequence by identifying the overlaps.
Other threads from the first post mention other advances from the HGP, of different sorts (technical, scientific, etc.). It's an interesting read, even though not about a deployment tool.
@zimoun @sharlatan @khinsen
=> More informations about this toot | More toots from Guillawme@fediscience.org This content has been proxied by September (3851b).Proxy Information
text/gemini