So I've changed a little bit the title actually, I've added, oh, non-coding RNAs in the story.

So we'll start with a general view of what is epigenetics. For those of you who don't know, essentially mathematicians, I guess, are not very familiar with the concept. So let's start with genetics.

Genetics. You of course know about Mandelian transmission and genetics, you heard about it this morning. Genetics allows a same phenotype with different genotypes, for example when you have a recessive inheritance of a trait, but genetics, if you take it basically, does not predict different

phenotypes with the same genotype. However, it's been known for a very long time actually, that there are many cases of identical genotypes and different phenotypes, which says that genetics cannot explain everything.

So I'll just give you a few examples of that. For example, the queen of the bees and the workers are generated from identical larvae, but they have very different phenotypes, as you can see here, and as you may know.

Another example which is closer to us is the cells from a single organism. All the cells from a complex organism come from a single cell. They have an identical genotype as far as we can say, but they come from the same cells

originally and then they become quite different when we are with an embryo, and you have completely different cells, brain, muscle, gut.

All these cells have the same genotype, but they have different phenotypes. This, so the question which has actually been on for a long time, but it's solved now, is how these cells can have the same genotype, but completely different phenotype, and the

answer comes from epigenetics. Now the modern definition of epigenetics is the analysis of stable and reversible changes in gene expression that do not involve modification of genetic information. And how can this happen at the molecular level?

So you know that genetic information is burned by DNA and it's located in the cell nucleus, in blue here in this picture, but DNA is not naked. In fact, in a human cell, in a human cell nucleus, there is, which has about 10 micrometer

of diameter, there is about 2 meters of DNA. So these very small nucleus cannot accommodate these 2 meters of DNA without compaction.

And in fact, DNA is compacted in the cells. I apologize for the resolution which comes from the change of computer, I guess. So you have the naked form of the DNA and then several degrees of compaction up to the

most compacted form, which is the mitotic chromosome. Now this compaction, in fact, is due to proteins, which are called the histone. So you have several of them, we are not going to go into the details here, but you

have two rounds of DNA, which is wrapped around the core of histone proteins and with a regular disposition. So these are called the nucleosomes. And the nucleosomes, so this is a chain of nucleosomes, and then this chain can be compacted

and then the new chain is compacted again, and so you have different levels of compaction. Now the problem is that in the compacted chromatin, the DNA is not accessible for protease cellular machineries, so that, and if you want to express a gene, you have

to have access to the DNA, to the RNA, the transcription machineries, the RNA

synthesizing machinery, so that in fact, compacted chromatin is linked to inactive gene, and open chromatin, decompacted chromatin, are potentially active genes. And indeed, if you look at the cell nucleus, you have different degree of compaction,

that you can even see by electronic microscopy, and you can distinguish what is called euchromatin, which is little compacted, and which includes active genes, from heterochromatin, which is highly compact, and which includes inactive genes.

Now chromatin compaction is controlled by chemical modifications of the chromatin. The DNA can be modified, it's methylated, and the histones, the proteins on which it is wrapped, can also be modified, they have little tails protruding out of the

nucleosome, which can be either acetylated, methylated, phosphorylated, and you have all sorts of different combinations of all these modifications of the histone tails. So that, and these modifications, these combinations of modifications are specific

of the different type of chromatin. In heterochromatin, the highly compacted chromatin, inactive chromatin, DNA is methylated, histones are deacetylated, and you have methylation on certain residues of the histone tail, for example, the Shui canine.

In contrast, when chromatin is active, in euchromatin, you have demethylated DNA, histones are acetylated, and they are methylated on other residues, such as H3E4, for example.

You can actually measure at the whole genome level these modifications, one by one, and this will give you the epigenome of the cell. Now if you take the genome as a book, the epigenome will tell you which are the open

pages, it's the epigenetic marks which will select the open pages of the book. So this, if you want, is the euchromatin, you can read it, this is the heterochromatin,

you cannot read it. So if we come back to the question of the cells which come from a single cell and have an identical phenotype, actually the difference is, at least in large part, due to the epigenome, because their epigenome is different.

So if you take a stem cell, for example, you will have the genes A, B, and C, which are methylated, etc., and compacted. Whereas if you take its daughter cell, the other genes will be methylated, compacted, etc.

So the pattern of genes which are open and the pattern of genes which are closed are different between these two cells. So the main question appears to be solved, we know what makes a different phenotype from a single genotype, but there are other open questions.

One of the questions which is very important actually is, can epigenome be modified by external signal? It's a socially important question because it includes the relationship between the genome and the environment, and you know all these problems created by the environment.

Well, we know that external signals can modify the epigenome. Like you have a ligand, for example, which will bind to a receptor, activates a chain of transduction, and eventually will change the chromatin very locally at very specific

loci from compacted to open and will activate the gene. At a more organism level, if we come back to the B model, it's in fact an external signal, which is the royal nectar with which the queen is fed, which makes the whole

difference between the worker and the queen. The fact that the queen is fed with this type of food, which modifies the epigenome of the larvae and makes a queen.

And it has been proven that some genes are methylated in the absence of royal nectar, and this will give a worker, whereas other genes will be activated by the royal nectar. And these genes control the size of the reactivity, etc., of the future bee.

In mammalian also, we have many examples of effect of an external signal on the epigenome. Here, for example, is the example of these mice, which are born from a stressed mother.

So when the mouse is pregnant, you stress it by noise, different noises or stuff like that. Now the mouse, which are born from a mother that was stressed during pregnancy, are themselves stressed, whereas the mouse, which is born from a well-treated mother, is not stressed

and it has been linked to the methylation of the gene, which is a gene which is expressed in the neurons and which actually defects, is associated with schizophrenia. And the mouse born from the stressed mother, this gene is methylated, whereas in the mouse

born from a well-treated mother, the gene is not methylated, so it's expressed. So it is clear, our environment, nutrition, external conditions affect the epigenome. Now another very important question, and this one is not solved at all,

is can epigenome modifications be transmitted to the next generations? Now if we take a somatic cell, like for example we have a culture, let's say fibroblast,

so differentiated cells, then all the daughter cells of this fibroblast will be fibroblasts, which clearly indicates that the epigenome is transmitted from the mother cell to the daughter cells, despite the fact, and I think it's a very interesting question actually, that during the mitosis

the chromosomes are highly condensed, it's the most condensed form of the chromosome. So, and despite this fact, they keep, they maintain this information, and actually we don't know how. But, now if we take, not the somatic cells, but the germ cells,

is the epigenome transmitted to the descendants? Well there are actually numerous studies in mouse and human suggesting that it might be the case, and if we go back to the stressed mice, actually the mice which are born from a stressed mother,

they are stressed, but they also have a good probability of giving themselves birth to stressed descendants, raising the possibility that this epigenomic status or feature has been transmitted.

And there are numerous studies like that in the literature. However this is a very controversial issue, as actually these experiments are very difficult to interpret. Is the effect due to molecular transmission of the epigenome to the embryo in the germ cells?

Or is it due to cultural factors and behavioral transmission? The stressed mother is stressed. She deals with, or it deals if it's a mouse, with her infants in a very specific way.

The metabolic factor is stressed, so there are particular hormones there. Exactly, it can be. Stress is chemical, don't you think? The stress is chemical, but the way that the children receive the stress, it will induce a chemical reaction in them as well.

But if the transmission is... What's an embryo for receiving these hormones? And who cares what you are receiving afterwards? An embryo in development was chemically modified... Yes, but if it gives birth, the same embryo will give birth... It keeps chemical signature, not behavior. Well we see that it cannot keep this chemical signature, at least not at the epigenomic level,

because if we move directly to the next slide, actually the epigenetic marks are erased during development. They are erased, almost completely erased. So you can always say that almost is the fact that it cannot be really.

So this is the... The chemistry in the blood of the mother, they are erased here, but in development, this chemistry of the mother... Yes, but that is okay for the first generation. No, no, but I see what Misha is saying. This is okay for the first generation, but it's not okay for the second

or sometimes the third generation. Okay, so at a very important stage, so here are for example the methylation marks, so you have the DNA from the mother and from the father, so it's methylated here, and then at a very important stage,

which is actually the stage where the embryo appears, and the appearance of the embryonic stem cells, then almost everything is erased. So the epigenetic marks are lost. And then they reappear and you have differentiation of the cells.

So how could the marks be transmitted if they are erased during the development? This is a completely unsolved issue at this point, but you can ask the question, are the epigenetic marks the only players or are they other players involved at the genomic level of course?

And good candidates are non-coding RNAs. Non-coding RNAs are transcribed from the non-coding genome. So it means that they don't have any open reading frame to make proteins. Previously, the non-coding genome was called junk DNA

because it was thought that it is absolutely useless. Now it's rather called the dark genome. Why? In fact, the complexity of organisms is not correlated with the gene numbers.

So if we take for example C. elegans, which has 90 or not far from 20,000 genes, it's a little worm, which has very little complexity, very little complexity, and it's a very small animal as well.

And if you compare that to human, human is thought to have about 24,000 genes, so you have very little differences in the number of genes. So the complexity of an organism is not linked to the number of genes.

However, it is inversely correlated with the percentage of coding sequences. The less coding sequences you have, the more complex the organism is. And in fact, in humans, it's 97% of the genome which is non-coding.

All the protein that you heard about this morning represents only 3% of the genome. The rest is non-coding. So if you take for example, this could be a human chromosome. Well, the coding part is here in red,

and all the rest is non-coding. So for long, this non-coding genome has been of no interest at all. Again, it was called the junk DNA. And then the discovery of small non-coding RNAs

triggered the interest for the non-coding genome. So the non-coding RNAs are the microRNAs that I'm sure you heard about many times. So they are... They've been discovered in C. elegans and they are very small RNA sequences

which bind to the messenger RNA and prevents the synthesis of protein, the translation of the messenger RNA. But small non-coding RNAs represent a very small part of the non-coding genome. They are very short and there are a few thousand of them

so they don't represent the whole thing. Now we know that actually almost all the genome, it's even more than 70% of the genome, is transcribed. We know that thanks to these new sequencing technologies

which allow to see molecules which are very little represented in a population because what we do in this technology is that we sequence one molecule, the molecules one by one. So if you do enough molecules, if you have a molecule which is very little represented, you can see it.

And because of course the whole, the transcription of the whole genome does not give rise to very frequent RNAs. It's been totally ignored up to now. But now we know that almost all the genome is transcribed

and it's mostly transcribed in long non-coding RNAs. And we actually know that long non-coding RNAs are able to modify the epigenome. It's been known for very long that there is a non-coding RNA which is called Xist

which covers one of the X chromosomes in somatic cells of mammalian females resulting in the epigenetic silencing of the anti-chromosome. You know that you have to have gene dosage on the X chromosome so that male and female have the same level of expression

of the genes which are in the X chromosome. So in nature there are different ways to achieve that. And in mammalian cells, one of the X chromosomes is inactivated. And it's inactivated thanks to this long non-coding RNA. Although the precise mechanism is not known.

Do you know that, is it known whether every single one of those bits of RNA are actually functional? No. Of course not. Not yet. How would you determine that?

Well, it's really starting now. It's been started for, I would say, five years now. What you do is, as usual, to analyze the function, you kill the molecule. No, but you have 70% of the genome. It's even more than that. You just said it's 70%, right?

So how are you going to walk through 70% of the genome and eliminate every single one? That's insane. Well, you will have to do that. Because there is no way that you can do a cell with only the 3% of DNA. How one would test that?

Well, what people are doing right now is that they are looking at specific long non-coding RNAs which are differentially expressed in this or that situation in healthy cells versus whatever. Whatever disease or cancer or whatever. And then they look into this differentially expressed.

I think it's a little bit like what was done with the genes at the beginning, you know? At the beginning, people looked at one gene, which is differently, blah blah, and we didn't have the whole genome sequence and blah blah. Now, if the question is, did someone imagine a way to prove that...

I mean, there are many different sequences here, right? How would you imagine doing the experiments? Yeah, so the question is, did someone imagine how to prove that the 70% are functional? I don't think anyone did at this point. Come on, we know that most of it is not functional

because most of it is not under constraint. Well, that's not right. I think that's not the right answer and I have a slide at the end to... Sure, sure, sure. No, no, no, no, no. We can have discussion. I don't think it's the right answer because if you take only the constraint, the evolution constraint,

then you will say that only the coding part is... And you will conclude that there is absolutely almost no difference between a mouse and a human because it's 99.9% the same genes. So you have to admit that a zone without constraint plays a role in this.

If you have in fact a big difference between a mouse and a human, biologically, I think it's gradual. You easily go on from our perspective, I think objectively, why too big a difference? Why is there a big difference between a mouse and a human? The difference is not big maybe, but it's the most important. We do not have whiskers.

No, no, but... I have the feeling it's a little bit more than that. Actually, Misha, you know that mouse is not a good model for most of the diseases because the mouse does not behave like the human. I mean, it's different, but it's not less or more complex. So to say more or less,

you need to have a measure of what you mean. Less what you mean more. The only point that can be said is that a mouse is different from a human, but if you take the genes, the enzymes and all that, they are very, very alike. The words say that it's not different.

The question is what we emphasize. Is appearance different? No, it's not only the appearance. Even biologically, the mouse is different. If you take, for example, the lifespan, a mouse lives two years. There's a number of parameters. Even if there are a thousand parameters, and different than 10. For example, one parameter is size.

Come on. It's not a big deal. You're being a small friend. Another parameter is the thinking that the mouse doesn't have at this point. Or maybe yes, but not the same type of thinking. Not the same type of thinking. We tend to overemphasize humanity. Our perspective is just the view.

That's my point. What you're saying is that we are animals and I think that's absolutely correct, but then there are many differences. And it's not only the mouse and the human but also, again, the C. elegans swarm. It's completely different. There are many, many genes which are in common.

Actually, the coding sequence of the genes, the mouse and human, are almost identical. That's actually a good point because what is very diverse is the non-coding region.

Exactly. It's my last slide, actually. It's my last slide. Let's move to the last slide. The one that could determine, theoretically, what is a mouse, what is a human. Exactly. That's my last slide. Maybe we can skip the last slide.

Okay, so what we can think is that maybe non-coding RNAs may keep the memory. They might be in the cytoplasm of the maternal DNA or they might be associated with the sperm DNA, whatever. But they can keep maybe the memory

of these epigenetic marks. And then we come to the last slide. So there are all the open questions. So non-coding RNAs represent 97% of the genome. They are thus a huge molecular region of how to explain yet unexplained functions. And there are many functions which are not explained by genes, simply by genes.

Also, and that's exactly the point of the discussion, in contrast to protein-coding genes, which may be very similar between the different species, non-coding RNAs are highly variable. They are not submitted to constraint from one species to the other

and from one individual to the other within the same species. They have the potential to play an essential role in the differences between species and in the difference between individuals within a species. Yes? I just want to talk to your point

about the coding regions being the same. That's like saying that in the human, the eye cells are really different from the cells on your hand, right? Like a difference between a mouse and a human. It has nothing to do with the different coding regions. It has to do with the way, how the coding regions are regulated, right? Yeah, yeah, but in total, I mean, yeah,

this difference within an organism, but in total, if you take the, it's like, I would say the, whatever, in total, if you take the human and the mouse, they are, and the genome in total, they are not identical, but very similar.

Yeah, but it's not about the actual sequence. It's about what is being, when is being transcribed when. Yeah, the timing of the expression might be different, but actually, the non-coding region can regulate the expression of the coding region, so because the non-coding region is different,

it could determine the difference in the expression. Of course, but my feeling, it's only a feeling of course, is that the non-coding region doesn't do only, I mean, right now, we see it through the eyes of the genetics and the coding sequences.

Everything again comes back to the coding sequences, okay? Everything is due to gene and protein. I don't think this is the case, and I have the feeling that the non-coding RNAs, they play a role in the, for example, in the brain functioning. So you started with the difference between different cells in the same organism,

with brain and whatever other cells got, and about that, we already, as biologists, we call it development, and we have the answers already in the genome, because we have the transcription factors and the enhancers, and we have, I'm not saying all, but we are beginning to,

yes, I just wanted to make a point. So we are beginning to understand it without looking at the non-coding. We know, at least one example exists, that it is functioning, so there will be more. Although we don't know how it works. We don't. No, but I mean, at least we know what it does. We are throwing at us 97% of the genome

to us biologists to figure out how we are going to figure out. You know, but the question is a good question, of course. Now, if we think, I don't think that the whole development is due to the genes, and if you take the example of the microRNAs,

they were absolutely unknown, and so we didn't even think that these things were working, so I wouldn't be so. Yes. So one point that we cannot, is it not completely true that they are not under constraints, because in different species, they keep the same position in some chromosomes

and seems to be related to these chromatin organization maps that are... No, the sequence is under less constraints. And there are differences in the non-coding genomes. There are parts which are more common or more conserved, and parts which are less,

but there are some, as if their chromosomal organization seems to be conserved across species, and there may be some functions at least for part of them to this chromatin. Yes, what you're saying, if I understand it properly, there can be common functions between species for the non-coding RNAs,

and what I'm adding is that might also be species-specific functions, or even individual specific functions. Yes. Let's suppose that you make an artificial genome where you get rid of all the everything, and you only have the open reading frames. Where would that... I don't think anyway, but maybe we should do it,

but I don't think that the cell would live. No, no, no, no, I didn't say that. I said that you get rid of these bits of... So you make only one small chromosome. You make one small chromosome with a coding sequence.

I'm actually very... No, but the way it's organized, the chromatin in particular shape, you need this mechanically, yes? No, no, no, no. So what I'm just saying is what he just said. You convert a human being into a yeast. But I'm not sure that human cell will live. How do you know that?

We have to try. We have to try. We have to try. Because the rest of the cell has been... No, because I think... I'm not saying that I'm right. Just what I think is that the rest of the cell has been designed... Not designed, but has evolved to work with the complex genome

and I'm not sure. I'm not sure that if you take... But it would be an interesting experiment to do. Anik, actually, I have a kind of provocative question. You explain us about epigenetic marks which connected with like chromatin

and now you say about non-coding RNAase and of course the third question, how you think you suggest the interplay between them, how this non-coding RNAase can be involved? Yeah, I didn't go into the details. Yeah, yeah, I didn't go into the details. But for example, the Xist... No, but the Xist RNA modifies... How Xist can we believe this?

The Xist RNA modifies the epigenetic marks on the chromosome, on the X chromosome, which is inactivated. And there are other examples of that. So the non-coding RNAase, they had the capacity to modify the epigenetic marks. But I don't think it's the only way they work. That's my point. Okay, let's go.

No, I need you to choose whom you want. You did not speak of plants. I think it has well been shown some couple of years ago already transmission up to eight or ten generation in Arabidopsis. Arabidopsis. I'm not a plant specialist, I have to say.

I'm very difficult. Yeah, no, no. What did Michel tell you? Epigenetic transmission. Yeah, yeah, yeah. It made a big sound in the literature. So I wanted to know what is since this...

But the problem is... Well, the problem in plants, I don't know if there is this erasement of the epigenetic marks. I'm sorry, I'm not a plant specialist. I don't have the answer. I know in Manal it appears. It occurs, but...

So a few comments. So first of all, in your experiment of getting rid of most of the genome and seeing what happens, that's been done in nature several times. So for example, the puffer fish has a genome which is 100 times smaller than other fish and humans. And it's still a fish, okay? So this proves that most of the non-coding genome

is not required to make a fish. Or this fish. So why do we study all this? Well... No, but I... Please, please. If I may comment, make me comment. If the evolution constraint is so strong and we don't need this 97% of genome,

why has it been kept during the evolution? This is population genetics. So genome size is not correlated to complexity in any way. Genome size is to do with the strength of natural selection. So species which have very small effective population sizes have very big genomes.

So plants, humans have a very small effective population site. We went through a bottleneck in history. This is why our genome is full of rubbish. This is very well understood. Bacteria, fungi have very effective, very large effective population sizes, very strong natural selection. Every single base pair is under selection and has a fitness cost when you do something.

Is it correct? Actually, it is not absolutely correct because coming back to plants, in plants it is well known, for example, if you cut half, at least half of non-coding genome of Arabidopsis, non-coding part...

Which one? The Arabidopsis part. It creates a normal plant from this story. But this is only in plants and maybe only in Arabidopsis. But this is very interesting. But maybe it is somehow connected with the case that in animals we have much less plasticity than in plants.

He just made the opposite point. Pardon? He just made the opposite point. He made a fish that is a natural experiment. You agree? I agree. I agree. Then I agree. But I don't know about animals. He is telling you about animals.

The same thing happens in animals. You have very different genome sizes for animals which are very similar to each other. So the genome size is not related to complexity. You don't need most of the genome. But you say about comparison and I say about experiment. It is not the same. Because comparison still doesn't say the actual result. Okay, but they agree.

They agree with each other. And the other thing is also this measuring which bits of the genome are important. So we can do this. One is by conservation during evolution. And this estimates that roughly 8% of the human genome is under constraint as in it has been under selection during primary evolution. And the other one now is

because we have so many sequences of individual human genomes we can measure the allele frequency, the frequency of the mutations at every base. So the human population is large enough such that every generation within saturation meets the genesis of the human genome now. So somebody is born with a mutation everywhere and through sequencing,

millions of genomes now, we know the frequency at which you get mutations in the polymorphism, within the population at each base. But when you say sequencing, the whole genome or only the exome? It's mostly the exomes but becoming the whole genome. And as this gets bigger and bigger then we will be able to measure exactly how important every base of the human genome is. So this is the other way to answer your experiment.

It's like we will have the numbers to do this to say this base in order to account the individual differences. So maybe this base is very important in general but this other base, which is only presenting two people, is the one base which makes the difference between these two people.

I don't understand where this 97% figure comes from. How do you know that 97%? I didn't do the experiment myself but it's the percentage of DNA which open reading frame of sufficient size

to make a protein. Because what I heard is that like 8% has been shown that has some... Well, it depends on the species. So in human it's 3% but in other species it's small. So it all depends on the species. No, but the 97% is definitely not all non-coding RNAs. This is your point. A lot of it is non-transcribed at all, you mean?

Yeah, of course. Well, at least there are parts of the genome that we cannot even sequence properly. Of course, we cannot know if they are transcribed or not. So we'll never come to 100% of things.

So why are you also focused on the genome? There are many more molecules which are passed from one generation to the next and I'm sure you've talked about germ plasm and glyco conjugated proteins and whatnot. I'm just curious to know. No, there might be part but at some point you have to remodel the genome because

if you take after the erasement of the marks at some point you recolor the marks and you start with a stem cell and so glyco conjugated they can induce the modification the epigenetic modification question. Yes, why not? But at this point it has not been proven that any sugar influences the genome.

I mean the structure of the genome. Did anybody ever do an experiment by microinjection of long non-coding RNAs? Yes, in fertilized eggs, no. And to show that they actually affect... Not yet. So how do we know they are important?

It's all correlation. It's an open question again. It's not a demonstrated fact. She wants us to do it. She is just suggesting an idea. Eric is not... She is just coming up with an idea. That's it. I totally agree that this is not proven.

But there are many examples of RNAs, non-coding RNAs, which are modifying the genome. The Xist is one. And there is another one which is very interesting. It's the paramecia. The paramecia has two nuclei, macronuclei and a macronuclei.

And the macronuclei has actually the whole genome. And it's very condensed. It's like the storage of the genome. And the other nucleus is very large. It has an condensed DNA and it's the part which is expressed. And during the division, the cell division,

the micronuclei will give rise to a micro and a macronuclei. So in the macronuclei, the sequences which are not expressed, are deleted. And this is directed by RNAs. So there are examples, many examples.

Sure. No, sure, of course. But it is exactly in paramecia that you have the heredity, non-DNA. Oh no, but yeah, but it's... Thomas on the bone. That's exactly his whole life. No, no, but I'm not talking about the way the...

I'm not talking about the genetics of the animals. I'm talking about the molecular... No, I'm talking about the mechanism. About something which is different. Yes, of course, but the molecular mechanism exists. That's what I'm saying, okay. Yeah, of course, in C. elegans also, yeah. Yeah, question, genetic.

But in a way, microRNA is not so well applicable now as an example of non-coding of these long non-coding RNAs, because they have their genes. They are transcribed as genes, but not protein genes. They're still not junk or black.

Yeah, but those are not protein gene either. It's not a protein gene, but they are not on other... So, but they have their... They are transcribed. They have their... There is a strong association between long-coding RNA and disease.

In disease, for example, in cancer. Yeah, yeah, yeah, absolutely. I believe in diabetes maybe as well, but... No, no, no, in cancer very well. There are several long... And these are those which are really looked at now. So 50% are coding from this long non-coding,

50% can code for micro proteins. Yes, that's right, but it has not been proven by these micro proteins. No, no, no. Some of these micro proteins are very important. Yeah, yeah, yeah. I'm aware that they're very small polypeptides, very small peptides actually.

So it's not proven that they have any function at this point, but it's, yeah, they are part of these long-coding. They have very small open reading frames. Are they synthesized? No, they are synthesized and they have a function. A function is not for sure, yeah.

They are synthesized, yeah. They are synthesized. Do you say all the peptides are very long, they are long-coding where they give digestion? Yeah, if they are very small, they can come from, yeah, yeah. Yeah, but one reason also why they are studied is because in different tissues, long RNAs have the same expression pattern.

So if you take a liver or a neurons, you see the same long RNAs expressed in the same tissues. So this is also one reason why they are, at a certain point, they try to understand what they do in the cell. Because there are kind of reproducible expression patterns

for these long RNAs in also in arthritic tissues. Well, but that's a correlation and also it can be caused because of the different types of transcription factors in the different tissues. Right, they seem to be under some regulations.

Okay, so we can take one last question. Do we have one? Everybody's hungry, I think. Everybody's hungry.