Now AI is designing antibodies too

Photo by Nataliya Vaitkevich on Pexels.com

Do you understand the human immune system, or the even vaster world of immunology? Neither do we. Too big, too vast, too complicated. And that’s despite 43 years of trying and working along side some pretty nifty immunologists, back in the day. But AI does, Fresh from its triumphs on protein design (LSS passim) its superior intelligence has now been turned on the knotty problem of how to run up new antibodies. Read this, AI designed antibodies race towards trials from the indefatigable Nature Briefing

Scientists say they are on the cusp of turning antibodies designed by artificial intelligence (AI) into potential therapies just a year after they debuted the first example of an entirely AI-designed antibody. Previously, the structure of antibodies proved somewhat of a black box to AI models. But new and improved models — such as an updated version of AlphaFold — have more successfully predicted the shape of flexible structures that give antibodies the specificity they need to bind to foreign molecules. Researchers at several companies now say they’ve designed ‘drug-like’ antibodies. Nature | 5 min read

There’s a lot to unpick here, and we won’t try to do it all. The first thing that stands out is how quickly this is moving from proof of concept to clinical reality, The second is that the possibility that designed antibodies will target receptor sites hitherto off-limits to their natural predecessors. Think infectious diseases first, and rightly. But the implications for cancer therapy -and dare we hope, neurodegenerative diseases?- are clear after a moment’s reflection. Designer antibodies will greatly reduce the need for much animal immunisation and testing. And, perhaps a best of all, a thriving commercial ecosystem of start-up companies is beginning to form around the new learning, ready to turn it into everyday reality in a hospital or medical prectice near you, gentle reader.

Although we can’t claim credit for these advances-you all know us too well- we think it striking that they have come from the qualities we prize . Careful observation. Null hypotheses testing. Slow steady work. And always looking for what proves you wrong not what proves you right. Those qualities are what save lives and generally make them better. A shame that they are being abandoned now by a hysterical ignorant population and so many of its foolish leaders.

#immunology #antibodies #medicine #health #cancer #disease #biotechnology

Stem Cell Therapy: Lamarckism by Force Majeure

First of all hats off to Oliver Chu, the brave boy from California who has just undergone a successful trial of stem cell therapy for a terrible condition called Hunters Syndrome. [1] as Ian Sample of the Guardian explains. It’s caused by a simple mutation in a gene called IDS-1 which controls the production of a vital enzyme Iduronate-2-sulfatase; without which the body cannot break down key sugars, leading to organ damage of all sorts and cognitive decline. The trick has been to extract the stem cells from Oliver’s blood: replace the faulty gene with a true copy using a viral vector; and pop them back in to Oliver, whenceforth they will thrive happily, self reproducing from their own line, and producing bountiful quantities of the enzyme for life.

And this for us is the key part. Let’s repeat : the new stem cells with the engineered gene will start their own self replicating line. In Oliver. Now Oliver himself started from a single stem cell-a single fertilised ovum, as do all living things. With DNA that was used to build every single following cell as it grew . An Ur stem cell if you like. But now. young Oliver has two. All the cells from his original cell, Plus the new line, from the engineered stem cell. whose line is now flooding his system with the good enzyme..

The central tenet of biology up to now is that we all of us-tigers, pterodactyls, humans, whatever-have a single unmodifiable line of DNA in our cells. Random variations may be passed to the next generation and tested by Natural Selection. But the actual DNA deep in the cells cannot be changed or modified. That’s the Darwinian positioned its held up pretty well for centuries. The alternative, proposed by Lamarck is that organisms are modified by the environment and this information can be learned inthe genes and passed on. So far there has been no evidence to support this view whatsoever . But what if the environment contains clever humans who can choose to modify DNA, and thereby create what are in effect hybrid organisms with two separate DNA lines-like young Oliver? Is this Darwinian? It’s not how it happens in nature, and its been done by force majeur. But it sounds a lot like Lamarckism from where we sit.

[1]https://www.theguardian.com/science/2025/nov/24/groundbreaking-uk-gene-therapy-manchester-hunter-syndrome

#stem cells #hunters syndrome #darwin #lamarck #evolution #medicine #health

PERT: Next step in gene editing offers real hope for hereditary diseases

Almost a quarter of hereditary diseases can be put down to mutations which break an established pattern of DNA, so it can no longer be read. No wonder they are called nonsense mutations. Often these mutations are expressed as STOP codons: just a short three letter sequence that stops protein synthesis dead, like a bad piece of coding in a computer programme. Now a new technique called PERT (Prime Editing RNA Therapy)allows the cellular process to override glitch in the DNA and resume synthesis. The new technique equips cells with engineered tRNAs that override these stop signals, letting the ribosome continue translation and produce the full protein. Here once again is Nature Briefing with one of their excellent short explanations Versatile gene-editing tool fixes nonsense, plus hyperlinks if you wish to delve deeper.

A multipurpose gene-editing tool can correct several genetic conditions in mice by restoring proteins that have been cut short by disease-causing mutations. The method, called PERT, uses engineered RNA molecules that allow protein synthesis to continue even when a DNA mutation tells it to stop prematurely. These ‘nonsense mutations’ comprise nearly one-quarter of known disease-causing DNA variants. As such, if PERT proves effective in humans, it could overcome the need to design bespoke treatments for individual diseases.Nature | 5 min read
Reference: Nature paper

There’s a lot to like here. Firstly the prime editing is straight out of the same stable as the CRISPR and Base Pair Editing techniques which we have heralded here for years (LSS passim) Secondly, unlike most gene therapies, which must be tailored to each mutation, PERT could treat many different diseases with a single editing agent. This is a huge shift in scalability. And if the suppressor tRNA is permanently installed in the patients genome, it is possible that only one treatment may be needed. Once again we are reporting at the early stages (that’s our brief) so all parties are careful to emphasise we are nowhere near clinical applications yet. However, just as we learned during the COVID 19 pandemic , the ability to intervene at the RNA level, precisely between gene and protein, appears to be one of the most fertile areas in medical knowledge for years to come.

#hereditary disease #RNA #DNA #molecular biology #health #medicine

Two good news blogs:#1 AI designer phages against antibiotic resistance

Its not often we bring you unabashed good news, gentle readers. Nor moreover, blow our own trumpet. But the following from Nature Briefing AI Helps design E. coli killing viruses not only unites so many of the themes we push here-(AI molecular design, multidisciplinary studies, bacteriophages etc etc)-that we think that the advance it represents it makes this one of our more significant blogs in months.

Using artificial intelligence (AI), researchers have designed novel viruses capable of killing strains of Escherichia coli. The team used the DNA of a simple bacteriophage called ΦX174 to guide AI models to generate viral genomes with the specific function of infecting antibiotic-resistant strains of E. coli. Researchers used the model’s suggested sequences to select 302 viable phages. When put to the test, 16 of these phages could infect E. coli, and combinations of them could kill three strains of the bacterium, a feat the original ΦX174 couldn’t pull off.Nature | 5 min read
Reference: bioRxiv preprint (not peer reviewed)

its certainly worth clicking on the Nature article and even the Preprint, which has a surprisingly well written summary

Old hands to this blog will recall our long standing worries about this organism. Normally Escherichia coli (named after the ingenious Dr Theodor Escherich) is a fine upstanding member of the microbiological community, being common in nature and a doyenne of experimental departments in microbiological schools. But certain strains of it are developing a profound resistance to our best antibiotics including piperacillin/tazobactam combinations. Which could have made it a very False Friend indeed. But now it seems that Dr King and his team have got ahead of the game.

Note the careful language, full admissions that peer review awaits, and generally understated claims that mark the true signs of trustworthy scholarship. How different from some situations where leaders of great nations go before the cameras and make huge unsubstantiated claims about phenomena of which they have no certain knowledge, But when you choose to believe only what you want to believe, fate has a nasty way of catching up eventually. Wait for the next blog and we’ll tell you more .

#E.coli #bacteriophages #AI #designer biochemistry #antibiotic resistance #microbiology #medicine #health

Mirror Organisms: the ultimate bioweapon?

Anyone who got beyond basic school science will recall the frustrating new level of complexity when the teacher first told you about stereoisometry. You recall-all biomolecules starting with the slightly complicated upwards really have two identical forms, left hand and right hand. Amino acids, proteins you name it. And life can only work with one. All amino acids in living things on this planet have left handed amino acids and right handed sugars. Of course living systems could work the other way round, It just has happened yet on this planet. Until now. Read this Debate heats up over mirror life from Nature Briefing

At a meeting this week in the United Kingdom, scientists are deliberating whether to restrict research that could eventually enable ‘mirror life’ — synthetic cells built from molecules that are mirror images of those found in the natural world. “Pretty much everybody agrees” that mirror-image cells would be “a bad thing”, says synthetic biologist John Glass. Such a cell might proliferate uncontrollably in the body or spread unchecked through the environment, because the body’s enzymes and immune system might not as readily recognize right-handed amino acids or left-handed DNA. But there are disagreements about where to set limits on research — the ability to evade degradation could also make such molecules useful as therapeutic drugs.Nature | 7 min read
Read more: Life scientist Ting Zhu, whose work explores various mirror-image molecular processes, considers how to bridge divergent views on such research. (Nature | 11 min read)

Unfortunately its the down size that worries us here, Not only the uncontrolled spread alluded to by the learned scientists above. But, as the world falls into the grip of authoritarian dictators and ever more powerful plutocrats, the potential these tools give them to get rid of surplus and redundant sections of humanity. Forever.

#isomers #biochemistry #bioweapons

LSS at 5:A blog of all our blogs

It’s funny, we’ve been doing this blog for more than five years now. And in response to growing numbers of readers and requests, we thought it might be time to provide a round up, not of the week, but of our whole outpourings which might be interesting to those who seem to have been trawling avidly through our archives of late.

It all started back in 2020, around the time of the great COVID-19 epidemic. Our initial aim was to raise awareness of the problem of antibiotic resistance in microbes, and the health dangers that posed. The idea was a short three paragraph hit the sort of thing that informed readers could take in over a quick coffee, while giving them a few links and references if they wanted to follow up. Just to keep it interesting, we started throwing in other topics on other areas of science. And these widened to include economics, social issues like women’s safety, and of course our regular Friday cocktail night, which certain readers still recall fondly.

Antibiotics and associated matters have remained well represented. We have looked for untapped sources in nature, even including the unlikely Komodo Dragon( LSS 3 5 21) the evolutionary arms race between bacteria and antibiotics which humans have been forced to join(LSS 8 6 23) and all sorts of new discoveries and techniques including AI (LSS 6 6 24) Being who we are, and untied to the constraints of any institution, we were quick to suggest that bacteriophages might be a useful adjunct to the general theme of overcoming resistant bacteria(LSS 17 3 22, 10 9 25 et al) Ever mindful that lack of antibiotics might not be the only catastrophe waiting we have provided handy little guides to what might happen if the magnetic poles flip, sea levels rise and even more endocrine disruptors are poured out from our factories. Other scientific tropes like evolution get a look in too. We enjoyed posing you a few puzzles on things like Homo naledi (LSS 4 4 21) the tools of Socotra (LSS 17 6 22) and even the possibility of Denisovan Fine Art( LSS 9 8 23) But these last were mainly for entertainment.

Our general theme has, we think been broad but consistent. The scientific method, of gathering objective evidence and analysing it by the rules of logic are the most reliable manner to fashion a passingly decent way of life. To this end you will have noticed is praise all kinds of people from journalists like Larry Elliott and Simon Kuper to more general thinkers like John Rawls, EO Wilson and Carl Sagan. We have tried to keep away from obvious stars like Darwin, Einstein, Bach, Keynes and the others as these thinkers speak for themselves. Instead we have tried to put forward slightly overlooked figures such as Ada Lovelace, Peter Ramus or Cassiodorus. Our Heroes of Learning feature is the place to look for those.

But above all we thank you, our readers, contributors and researchers for all their good companionship. All those who posts likes, shares and comments-it shows someone out there is interested. We wish all of you well with your various blogs, careers, lives and families. As Gore Vidal observed , it is the top one or two percent who carry knowledge through and pass it from generation to generation. And you are in it.

#antibiotic resistance #bacteriophages #environment #pollution #economics #history #evolution #science #reason #cocktails

CRISPR gallops ahead (article contains a warning for xenophobes)

Warning: this article may make uncomfortable reading for xenophobes everywhere)

Progress in CRISPR-Cas-9 (Clustered Regularly Interspaced Short Palindromic Repeats)[1] and the associated enzyme is getting faster and faster. We started reporting on this truly innovative technique in 2020 and regular readers will recall updates ever since. Only four years ago it still felt a bit theoretical. But now radical applications are coming thick and fast Read this from Nature Briefing CRISPR horses spark debate reporting on the rather recondite world of polo pony breeding

the horses pictured above{*} are the first of their species to have been created with the help of the CRISPR–Cas9 gene-editing technique. They are clones of the prize-winning steed Polo Pureza, with a tweak to myostatin — a gene involved in regulating muscle development — that is designed to quicken their pace. Critics say that genetic manipulation has no place among polo’s traditional breeding practices — it has already been banned by some of the sport’s governing bodies. But a zoo of CRISPR-edited animals, from cows to sheep, is gaining acceptance in agriculture.Nature | 5 min read

^{{*} sorry LSS readers-we can’t show this}^-ed

In one sense there’s nothing new here. Humans have been modifying the genetics of plant and animal species since the dawn of the Neolithic. CRISPR and other base editing techniques have simply speeded the whole process up by making specific, designed changes and crucial nodes in the subject organism’s development. There is every reason to suppose that any number of new modifications to animals(and crop plants such as wheat) will be developed in the next few years. Some may even enable us the preserve the integrity of food supplies despite the ravages of things like plastics pollution and global warming. Also, as we have also reported here, gene editing is beginning to show real applications in medical fields such as sickle cell disease and certain cancer therapies. All of which leads us to an intriguing thought.

If ponies may be so easily modified, why not humans? One could start small by just modifying athletes and other small groups. Yet eventually the techniques could become ubiquitous in our species. Hang on-our species? Because the genetic differences between beings consisting entirely of CRISPR modified genes and the rest of us would then be far, far greater than those currently existing between our different races and ethnic groups. Are xenophobes everywhere already wasting their own time?

[1]https://www.yourgenome.org/theme/what-is-crispr-cas9/

#CRISPR Cas 9 #base pair #medicine #biotechnology #sickle cell #agriculture #stock breeding

How life evolved long ago is absolutely relevant today

Long suffering readers of this blog will recall our occasional sallies into the remote past. Like some latter day Doug McClure we occasionally take you into a world stuffed with dinosaurs, ape men and pterodactyls, to the detriment of more relevant stuff on antibiotics or the US Ten Year Bond. And so, although we were privately raving about this piece below called How did life get multicellular? from Nature Briefing, we thought we ‘d spare you from our private obsessions about things that took place between 800 -600 million years ago.

Until a chance encounter with one of more intelligent friends in the car park at our Spanish Conversation group produced the most inspiring thought. “All those Choanoflagellates. protometazoans. Filasterea. whatever, have to do several things if they are to succeed in living together. To glue up to each other. To signal little messages. To co-ordinate the cycles of cell division. Just like cancer cells have to, in fact. And then it hit us. These funny little organisms are the perfect way to model the behaviour of cancer cells. Not just the molecular and genetic mechanisms, but also the Information and Complexity models we must build to understand them: a cancer cell is a typical metazoan cell gone wrong.

Which confirmed a very old principle of this blog. All research however abstruse it may seem, will have a pay off somewhere one day. If it doesn’t benefit the economy, it will make us live longer; sometimes it may do both. These researchers are not just having fun on the edge of time: they may be contributing directly to the study of a disease which will kill half of us. There’s a thought for anyone who wants to cut university budgets or meddle with the findings of scientists.

To play out we shall first post the Nature Briefing paragraph. If you can get past that we’ve some supporting evidence for our basic proposition. We hope both will inform

Across all forms of life, the move from being single-celled to multicellular seems to have happened dozens of times — for animals, though, the jump was one-and-done. The unique cocktail of environmental and genetic factors that helped animal ancestors make that jump still eludes our understanding. To investigate, researchers are focussing on unicellular organisms that ‘dabble’ in multicellularity, occasionally forming colonies of many cells. By studying these organisms as they flit between the two states, scientists are hoping to illuminate how multicellularity stuck in animals — and what sparked the single successful event that gave rise to the animal kingdom.Nature | 11 min read

ASTRACT BECOMES APPLIED

This work discusses how cancer disrupts the gene regulatory networks (GRNs) that evolved to coordinate multicellular life. These networks balance genes inherited from unicellular ancestors (handling basics like metabolism and division) with newer multicellular genes (handling coordination, differentiation, and tissue integrity). https://genomebiology.biomedcentral.com/articles/10.1186/s13059-024-03247-1

and this how somatic mutations in early metazoan genes specifically disrupt the regulatory links between unicellular and multicellular gene networks. The result? Tumours behave like rogue unicellular entities, ignoring the cooperative rules of multicellularity. Some of these disrupted genes even correlate with drug response, hinting at therapeutic relevance

thanks to R Muggridge

https://elifesciences.org/articles/40947

#cancer #evolution #multicellularity #medicine #health #choanoflagellates

Bacteriophages v Bacteria: this arms race offers opportunities

We’ve always hymned the praises of bacteriophages here (LSS passim): that they will be a vital second option to supplement the next generation of antibiotic drugs. But we have a confession. We didn’t understand them. We didn’t appreciate that they are biological systems (viruses) interacting with other biological systems (bacteria). And as such, will obey all the usual rules of all such systems, such as arms races between predator and prey, Now a new article by Franklin Nobrega for the Conversation puts that right. [1]

Bacteria have evolved some fascinating defence mechanisms to ward off the relentless attacks of their phage enemies. These involve cutting the nuclear material of the viruses: building up strong cell walls and cellular shutdown mechanisms which act a bit like your IT Department does when it detects a global virus attack on your building’s systems. Recently Franklin and his team have investigated an early warning system called KIWA which gives the bacteria advanced notice that an attack is imminent. To which phages have in turn responded by their own mutations, and so it goes on, etc etc.

There’s a lot to encourage us here. Firstly, human knowledge of bacteriophages and their ways is deepening all the time, always a good thing. In fact Franklin is part of the University of Southampton phage collection project which we showcased here a few weeks ago (LSS 1 7 25) More strikingly, as two systems attack each other in an arms race, they leave little gaps, tiny vulnerabilities, which outsiders can exploit. The promise of new drugs and new bioengineering techniques looks very real indeed. Especially, we suggest if information scientists and complexity theorists are brought in to work alongside the biological teams. All in all, a rather good day for those of us interested in the problems of microbial antibiotic resistance. Go boldly, gentle readers, and be of good cheer.

[1]https://theconversation.com/how-ancient-viruses-could-help-fight-antibiotic-resistance-261970?utm_medium=email&utm_campaign=Latest%20from%20The%20Conversation%20

#microbial antibiotic resistance #bacteria #bacteriophage #health #medicine #phage collection project

This is what awesome intelligence looks like

No we’re not writing about ourselves as some hardened readers may have already guessed. Because we couldn’t achieve what the researchers in these two stories, both from Nature Briefing, have indeed achieved. Sorry guv- we don’t have their intellectual bandwidth . Cognitive intricacy. Brains, in good old fashioned English But we know intelligence when we see it, and we know you do too.

Algorithms design remarkable enzymes Researchers have used computer algorithms to design highly efficient synthetic enzymes from scratch, reducing the number of tedious hands-on experiments needed to perfect them. The products facilitate a chemical reaction that no known natural protein can, with an efficiency similar to that typically achieved by naturally occurring enzymes. One design was also 100 times more efficient than similar enzymes previously crafted using artificial intelligence. In comparison to enzymes that occur in nature, the algorithm’s creations are less complex and can’t grapple with multi-step chemical reactions, but they’re proof that the approach has promise.Nature | 4 min read
Reference: Nature paper

The medical potential of designer enzymes will not be lost on readers as intelligent as our own. Yet some discoveries are to be relished not for their use, but for what they tell us about the world and our real place in it Try this for size

Dragon Man was a Denisovan Ancient proteins and mitochondrial DNA extracted from the ‘Dragon Man’ fossil — a cranium found in northeastern China that is at least 146,000 years old — have confirmed that it belonged to a Denisovan, an archaic human group. The fossil is the first skull to be definitively linked to the group, which sheds light on what the ancient people looked like, putting an end to decade-long speculation.Nature | 5 min read
Reference: Cell paper & Science paper

Learning. Reason. Curiosity. Handy, aren’t they? Their absence can lead to different outcomes indeed. As the inhabitants of certain regions of the globe know only too well.

#protein #mRNA #medicine #health #evolution