How far away are we from decoding the entire human genome?

[fredreload]

I understand some people have their genome fully sequenced, but that doesn't mean we understand what the genome does. So when is it that we will finally be able to know that a particular sequence represents height or something else having the genome fully decoded?

[CharonY]

Pretty far. In fact, we do not have clear approaches to that end. Or rather, we have figured out a lot of things that don't work well over the last 10-15 yearxps. So any prediction would be a wild guess.

[fredreload]

Pretty far. In fact, we do not have clear approaches to that end. Or rather, we have figured out a lot of things that don't work well over the last 10-15 yearxps. So any prediction would be a wild guess.

What I can think of is to observe the earlier stage of human development, zygote->blastocyst->body forming using a non-intrusive laser scanner. The thing is I'm not sure if this would be ethical, you'll need external birth for this. Then you get a 3D view of all the molecular mechanisms involved. Once you get this you can replay it however many times you want, I think, theoretically. We can't do it yet because there isn't external birth and I doubt anyone is going to volunteer for this

Edited September 19, 2016 by fredreload

[Delta1212]

You make that sound significantly easier than it would actually be to do.

[fredreload]

You make that sound significantly easier than it would actually be to do.

Well, laser scanning is a powerful technique, but to observe the DNA might require laser spectroscopy or using dyes. I am not sure about the resolution for such a technique not to mention observing a specimen in vitro. The hard part is to get the observation technique, after that we can probably start with some smaller specimen

 

P.S. The only thing I could find is hyper spectral imaging, with some shading applied it could be quite convincing

Edited September 19, 2016 by fredreload

[CharonY]

None of these have direct impact on genomic research.

[Arete]

As an evolutionary biologist, my first question would be which human genome?

 

Human genomes accumulate around 64 point mutations on average per generation, and vary between individuals and populations - genomes are not typological, so ultimately you're looking at a moving, changing target. not a singular entity.

[CharonY]

As a molecular biologist I usually takes samples from the slowest student. .

[Arete]

Bacteria complain less and their genomes take up less of my hard drive. Plus, the ethics committee doesn't care about them

[fredreload]

Well the point is from zygote to human takes cellular divisions and as it grows the gene expression is constantly changing, which part of the DNA shows which part of the body is being developed. And yes there would be microRNA involved as well as how it is instructed to perform the task. Nevertheless, once the human shape takes place I feel that it would be harder for us to find the DNA that instructs how cell should be differentiated to take the human form. For a grown up human the shape doesn't change anymore, so I think less can be observed in comparison to the developmental stage. Unless you have regeneration property and you observe how the body parts regenerate with microRNA

[Arete]

Well the point is from zygote to human takes cellular divisions and as it grows the gene expression is constantly changing, which part of the DNA shows which part of the body is being developed. And yes there would be microRNA involved as well as how it is instructed to perform the task. Nevertheless, once the human shape takes place I feel that it would be harder for us to find the DNA that instructs how cell should be differentiated to take the human form. For a grown up human the shape doesn't change anymore, so I think less can be observed in comparison to the developmental stage. Unless you have regeneration property and you observe how the body parts regenerate with microRNA

 

Cell differentiation is but one of the many functions of the genome. Even if you completely understand the genetic underpinnings of differentiation (and we do have a reasonable understanding of it) you'd still be a long way from a complete understanding of the whole genome... then the damn thing mutates and does something new and you have more work to do.

[CharonY]

Well the point is from zygote to human takes cellular divisions and as it grows the gene expression is constantly changing, which part of the DNA shows which part of the body is being developed. And yes there would be microRNA involved as well as how it is instructed to perform the task. Nevertheless, once the human shape takes place I feel that it would be harder for us to find the DNA that instructs how cell should be differentiated to take the human form. For a grown up human the shape doesn't change anymore, so I think less can be observed in comparison to the developmental stage. Unless you have regeneration property and you observe how the body parts regenerate with microRNA

 

Let me give a bit perspective on functional genomics from the bacterial side. Bacteria are unicellular and generally do not have complex cell/tissue differentiation to speak of. Moreover, you can manipulate most quite easily to look at gene functions. Now, the best characterized bacterium (and by extension, organism) is Escherichia coli K12 (clonal line, i.e. also ignoring mutational variants), which has been sequenced ~ 20 years ago (I think, gosh I am old). Since then we have poked it in any way we can, mutated basically every (non-essential) gene, profiled with proteomics, metabolomics and so on and so forth. Still, for about 25% of the genome we do not have a clue what the genes are doing. And if we are honest, for at least 50% the predictions are rather vague. So even in a simple, best understood organism we have a huge knowledge gap that we cannot fill by merely adding observatory analyses. You can imagine the added complications for multicellular organisms.

[Delta1212]

 

Let me give a bit perspective on functional genomics from the bacterial side. Bacteria are unicellular and generally do not have complex cell/tissue differentiation to speak of. Moreover, you can manipulate most quite easily to look at gene functions. Now, the best characterized bacterium (and by extension, organism) is Escherichia coli K12 (clonal line, i.e. also ignoring mutational variants), which has been sequenced ~ 20 years ago (I think, gosh I am old). Since then we have poked it in any way we can, mutated basically every (non-essential) gene, profiled with proteomics, metabolomics and so on and so forth. Still, for about 25% of the genome we do not have a clue what the genes are doing. And if we are honest, for at least 50% the predictions are rather vague. So even in a simple, best understood organism we have a huge knowledge gap that we cannot fill by merely adding observatory analyses. You can imagine the added complications for multicellular organisms.

And that's before you even throw in ethics as a complicating factor. You can't just keep cloning humans with different genes turned off to see what breaks.

[CharonY]

You could, however, use any other animal (and have it approved by the ethics committee), after all, genetically the difference between mammals is not that large. And even with recent advances in developmental biology, we remain largely clueless. We do have a qualitative grasp on some major signaling cascades and developmental switches, but the vast majority is still a black box. Looking at the parts is the easy bit, figuring out what they do and how they interact with each other is another entirely.

But again, we are talking about writing a literary masterpiece using a language of which we basically just roughly know what the letters are. Sure, we could continue to analyze the composition and frequency of the letters, but that does not automatically tells us what the words are, what they represent and even less about grammar. Then we add to the pile that the letter composition is dynamic and constantly changing.

[fredreload]

Hmm, you got 4 neucleotides and around 20 amino acids, from these number you can predict just about all the instructions given to a cell unless you are doing something like 20 factorial. I always thought you'd be able to track the DNA code, how it becomes mRNA and eventually protein in the ribosome based on these amino acids. As to how the instructions differ for a bacterium, I can't really say, I'd trace the mRNA, amino acid, and how it programs the protein to react. As for how DNA works for cell division like from zygote to blastocyst you'll have to check the gene expression and essentially the instruction that signal human shape to take place and differentiation of tissues. I always felt like this is kept secret or that my understanding of Biology isn't that great but clearly each cell has a different gene expression and something is coordinating these gene expressions to form the human shape and instruct cell differentiation. Arete probably has a better understanding of this

 

P.S. Well but, I bet you are not examining the bacteria at a molecular level->hyperspectral imaging

Edited September 20, 2016 by fredreload

[CharonY]

You are still looking at the wrong/easy thing. We have a somewhat decent grasp to look at molecular composition (at least if we ignore dynamics for the most part). But the problem is that changes in expression tells you little what they actually do. The reason being that they are not activated exclusively by some event, but many are induced by a multitude of intra-and extracellular signals. And your repeated claims of imaging still does not make any sense as you do not add functional information to the mix.

[fredreload]

You are still looking at the wrong/easy thing. We have a somewhat decent grasp to look at molecular composition (at least if we ignore dynamics for the most part). But the problem is that changes in expression tells you little what they actually do. The reason being that they are not activated exclusively by some event, but many are induced by a multitude of intra-and extracellular signals. And your repeated claims of imaging still does not make any sense as you do not add functional information to the mix.

Well, my idea is about observing everything from DNA to mRNA, what gene expression is is the rate of a certain strand of DNA copied to the mRNA, if it is not expressed it is not copied. We agree that this messenger RNA is passed to the bacteria ribosome and eventually synthesize the protein there, then we trace this protein to see what gets developed all using hyperspectral imaging at real time like an animation. Sounds like science fiction, well at this point it might be. And if simply synthesizing the protein with messenger RNA in the ribosome is not the whole picture then you can explain to me what is missing. Well this is all theoretical, I respect you for doing the actual work on the bacteria and you are free to tell me what I should be looking for

Edited September 20, 2016 by fredreload

[Arete]

Well, my idea is about observing everything from DNA to mRNA, what gene expression is is the rate of a certain strand of DNA copied to the mRNA, if it is not expressed it is not copied. We agree that this messenger RNA is passed to the bacteria ribosome and eventually synthesize the protein there, then we trace this protein to see what gets developed all using hyperspectral imaging at real time like an animation. Sounds like science fiction, well at this point it might be. And if simply synthesizing the protein with messenger RNA in the ribosome is not the whole picture then you can explain to me what is missing. Well this is all theoretical, I respect you for doing the actual work on the bacteria and you are free to tell me what I should be looking for

 

 

We can use RNAseq do examine gene expression already. Also, knowing the sequence of amino acids in a protein doesn't mean you know its 3D structure (assuming it's fixed, as many proteins fold differently in different environments - hence the environment plays a role in how the protein functions) , so no, simply knowing a particular amino acid sequence is expressed does not mean you know what the protein looks like, what it does, or how different variants of the gene result in differently functioning proteins.

 

Furthermore, plenty of functional genomic regions are non-coding, regulatory regions. Expression/translation of these genes doesn't necessarily tell you about their function.

[fredreload]

 

 

We can use RNAseq do examine gene expression already. Also, knowing the sequence of amino acids in a protein doesn't mean you know its 3D structure (assuming it's fixed, as many proteins fold differently in different environments - hence the environment plays a role in how the protein functions) , so no, simply knowing a particular amino acid sequence is expressed does not mean you know what the protein looks like, what it does, or how different variants of the gene result in differently functioning proteins.

 

Furthermore, plenty of functional genomic regions are non-coding, regulatory regions. Expression/translation of these genes doesn't necessarily tell you about their function.

Right well I'm thinking of creating an animation that looks like this(man I feel like a kid asking for candy), I'm not sure how accurate this animation is. The amino acid sequence is not the only thing observed, but how it folds into a 3D shape and how it's used as well as its functions the whole picture. You can let me know if that is not possible with the current technology

[CharonY]

Even if you had the structure (e.g. via crystallographic approaches), figuring out their function is still a long. loooong process. Or take proteomics, we routinely look at hundreds or thousands of proteins and monitor their expression under many conditions. Still, we generally rely on our old-school knowledge of known biochemical pathways to make sense out of the patterns. Again, what is missing is not the patterns, but figuring out what they mean. Even if you know the general role of a protein (say, vesicle trafficking) it will fulfill many different roles depending on what they particular cells needs to be done. E.g. it could direct signaling vesicles, or nutrients, or direct lipid genesis etc. Now you have thousands of them of which you vaguely know they roles or not at all.

It requires extensive manipulation and many many experiments even on simple cells to use that information to tease out potential functions.

 

Edit: the movie is a cartoon we have no way of getting that. It is one of the grails in structural biology, where we use e.g. femtosecond X-ray crystallography to try to capture confirmational changes. There are only a handful of examples where we might have seen something, and it took years in each case (not to mention a massive X-ray source). And this is only for minute changes (say, photoactivation of a molecule), nothing like a complex folding.

 

But to re-iterate, while structure can help in figuring out function, it is a non-trivial process. And even if you know that that enzyme specifically binds a certain substrate, you still do not know the physiological role. I.e. when does it do that, to what reason, and what are the consequences if does not do that. The reason being that all the proteins are connected via massive networks. If you poke a whole in one area, it may be caught by something else entirely, so you do not see a phsyiological reaction. Or the opposite may happen, where it actually affects a distal part of the network, which leads to entirely unexpected results. Again, the limiting factor are not the measurements as such, but good theoretical framework or models which helps us to understand these vast amount of data. This is where bioinfomratics originated. And while there have been approaches (such as in the biochemcial modeling frameworks) it is still very limited and barely works on the single cell level. Anything beyond that tend to be empirical models (if at alll).

Edited September 20, 2016 by CharonY

[fredreload]

Even if you had the structure (e.g. via crystallographic approaches), figuring out their function is still a long. loooong process. Or take proteomics, we routinely look at hundreds or thousands of proteins and monitor their expression under many conditions. Still, we generally rely on our old-school knowledge of known biochemical pathways to make sense out of the patterns. Again, what is missing is not the patterns, but figuring out what they mean. Even if you know the general role of a protein (say, vesicle trafficking) it will fulfill many different roles depending on what they particular cells needs to be done. E.g. it could direct signaling vesicles, or nutrients, or direct lipid genesis etc. Now you have thousands of them of which you vaguely know they roles or not at all.

It requires extensive manipulation and many many experiments even on simple cells to use that information to tease out potential functions.

 

Edit: the movie is a cartoon we have no way of getting that. It is one of the grails in structural biology, where we use e.g. femtosecond X-ray crystallography to try to capture confirmational changes. There are only a handful of examples where we might have seen something, and it took years in each case (not to mention a massive X-ray source). And this is only for minute changes (say, photoactivation of a molecule), nothing like a complex folding.

 

But to re-iterate, while structure can help in figuring out function, it is a non-trivial process. And even if you know that that enzyme specifically binds a certain substrate, you still do not know the physiological role. I.e. when does it do that, to what reason, and what are the consequences if does not do that. The reason being that all the proteins are connected via massive networks. If you poke a whole in one area, it may be caught by something else entirely, so you do not see a phsyiological reaction. Or the opposite may happen, where it actually affects a distal part of the network, which leads to entirely unexpected results. Again, the limiting factor are not the measurements as such, but good theoretical framework or models which helps us to understand these vast amount of data. This is where bioinfomratics originated. And while there have been approaches (such as in the biochemcial modeling frameworks) it is still very limited and barely works on the single cell level. Anything beyond that tend to be empirical models (if at alll).

Right, I was looking into the attosecond laser and its detection on molecular changes before I look into laser spectroscopy. I guess I had my hopes up thinking the whole thing can be visualized

 

P.S. I thought it would be possible to track each protein using a visual analysis program on the animation, but you say acquiring such an animation is not possible at this point then I guess we'll have to wait and see

 

 

The part about the visualization technique 0:22~0:40, year 2030

Edited September 20, 2016 by fredreload