5 ms·

I don't see in the article why this is useful or what is it used for?

by neom 5y ago

madhato 5y ago
Its simple really. First you simulate a single cell, then a sperm and an egg cell. Then you simulate a virtual a world of virtual captive humans to do our work for us without payment.
joejoesvk 5y ago
i don't even know where to start. you could simulate one cell, then two cells then 4..suddenly you could have an organisms...hell you could see organism that could have lived on earth. maybe it'll one day help with cancer research.
- neom 5y ago
  Oh I see, I didn't realize it extrapolated out that way, this was a good explanation that makes a lot of sense to me, thanks. Also.. that's f'ing cool.
JabavuAdams 5y ago
It's generally not possible to see where everything is in an actual cell, in realtime, due to the sizes of the components. So most of molecular biology relies on very clever lab techniques to indirectly infer what cells are making and doing. Cells are like little cities in terms of the complexity of their biochemistry. We want to ask questions like "How does this cell respond to this chemical/drug/change in environment." Imagine trying to understand in detail the gun crime epidemic in a city, if you can only see objects larger than 100 m on a side. You wouldn't see people, cars, or many buildings. We want to be able to understand, explain, predict, and control cellular process, but so far we have to be quite indirect. Understanding these things at a mechanistic level, in realtime would revolutionize our ability to understand, repair, and build biological systems.
JabavuAdams 5y ago
For instance, the cartoon version of DNA that is presented to even lower-year biology undergraduates is of linearized strands. But of course, it's really all spooled and tangled and crunched up into the nucleus of cells. Note that the cell they simulated is of a prokaryote (no nucleus, much simpler cellular processes than e.g. a mammal cell). About 1-2% of our genes make proteins, although the proportion is much larger in single-celled organisms (less redundancy in the genome, no splicing, etc.) So when you hear that e.g. genes turn on or off, this is not a switch. It's literally some sections of DNA being unwound, and large complexes of mutually interacting molecules probabilistically glomming on and off the DNA. The actual 3d layout of this DNA "ramen" matters to e.g. bring promoter regions of genes close to the actual genes they control. So basically, we have a schematic-level understanding of cellular processes, but to see the actual 3D interactions in realtime would be extremely illuminating.
- JabavuAdams 5y ago
  I should say that this work is not simulating things at this detail. Instead, it's more like a biophysical model of a bunch of chemical reactions with rate information. It probably boils down to a big system of coupled differential equations, at different timescales. So, it's a statistical level of detail, but still very informative.

I don't see in the article why this is useful or what is it used for?

by neom 5y ago

I don't see in the article why this is useful or what is it used for?

Its simple really. First you simulate a single cell, then a sperm and an egg cell. Then you simulate a virtual a world of virtual captive humans to do our work for us without payment.

joejoesvk 5y ago

i don't even know where to start. you could simulate one cell, then two cells then 4..suddenly you could have an organisms...hell you could see organism that could have lived on earth. maybe it'll one day help with cancer research.

neom 5y ago
Oh I see, I didn't realize it extrapolated out that way, this was a good explanation that makes a lot of sense to me, thanks. Also.. that's f'ing cool.

JabavuAdams 5y ago

It's generally not possible to see where everything is in an actual cell, in realtime, due to the sizes of the components. So most of molecular biology relies on very clever lab techniques to indirectly infer what cells are making and doing. Cells are like little cities in terms of the complexity of their biochemistry. We want to ask questions like "How does this cell respond to this chemical/drug/change in environment." Imagine trying to understand in detail the gun crime epidemic in a city, if you can only see objects larger than 100 m on a side. You wouldn't see people, cars, or many buildings. We want to be able to understand, explain, predict, and control cellular process, but so far we have to be quite indirect. Understanding these things at a mechanistic level, in realtime would revolutionize our ability to understand, repair, and build biological systems.

JabavuAdams 5y ago

For instance, the cartoon version of DNA that is presented to even lower-year biology undergraduates is of linearized strands. But of course, it's really all spooled and tangled and crunched up into the nucleus of cells. Note that the cell they simulated is of a prokaryote (no nucleus, much simpler cellular processes than e.g. a mammal cell). About 1-2% of our genes make proteins, although the proportion is much larger in single-celled organisms (less redundancy in the genome, no splicing, etc.) So when you hear that e.g. genes turn on or off, this is not a switch. It's literally some sections of DNA being unwound, and large complexes of mutually interacting molecules probabilistically glomming on and off the DNA. The actual 3d layout of this DNA "ramen" matters to e.g. bring promoter regions of genes close to the actual genes they control. So basically, we have a schematic-level understanding of cellular processes, but to see the actual 3D interactions in realtime would be extremely illuminating.

JabavuAdams 5y ago
I should say that this work is not simulating things at this detail. Instead, it's more like a biophysical model of a bunch of chemical reactions with rate information. It probably boils down to a big system of coupled differential equations, at different timescales. So, it's a statistical level of detail, but still very informative.