= Molecular biology technologies
See also: https://github.com/cirosantilli/awesome-whole-cell-simulation
As of 2019, <Moore's law>[the silicon industry is ending], and <molecular biology> technology is one of the most promising and growing <deep tech>[field of engineering].
\Image[https://web.archive.org/web/20191008120152im_/https://www.karlrupp.net/wp-content/uploads/2018/02/42-years-processor-trend-625x396.png]
{title=42 years of <microprocessor> trend data by Karl Rupp}
{description=Only transistor count increases, which also pushes core counts up. But what you gonna do when atomic limits are reached? The separation between two silicon atoms is https://www1.columbia.edu/sec/itc/ee/test2/pdf%20files/silicon%20basics.pdf[0.23nm] and 2019 technology https://en.wikipedia.org/wiki/5_nanometer[is at 5nm] scale.}
{source=https://www.karlrupp.net/2018/02/42-years-of-microprocessor-trend-data/}
Such advances could one day lead to both <unconditional basic income>[biological super-AGI and immortality].
<Ciro Santilli> is especially excited about <DNA>-related technologies, because DNA is the centerpiece of biology, and it is programmable.
First, during the 2000's, the cost of <DNA sequencing> fell to about 1000 USD per genome in the end of the 2010's: <image Cost per genome vs Moore's law from 2000 to 2019>{full}, largely due to <Illumina>["Illumina's" technology].
The medical consequences of this revolution are still trickling down towards medical applications of 2019, inevitably, but somewhat slowly due to tight privacy control of medical records.
\Image[https://upload.wikimedia.org/wikipedia/commons/thumb/0/01/Cost_per_Genome.png/1024px-Cost_per_Genome.png]
{title=Cost per genome vs Moore's law from 2000 to 2019}
<Ciro Santilli> predicts that when the 100 dollar mark is reached, \i[every] person of the https://en.wikipedia.org/wiki/First_World[First world] will have their genome sequenced, and then medical applications will be closer at hand than ever.
But even 100 dollars is not enough. Sequencing power is like computing power: humankind can \i[never] have enough. Sequencing is not a one per person thing. For example, as of 2019 tumors are already being sequenced to help understand and treat them, and scientists/doctors will sequence as many tumor cells as budget allows.
Then, in the 2010's, https://en.wikipedia.org/wiki/CRISPR_gene_editing[CRISPR/Cas9 gene editing] started opening up the way to actually modifying the genome that we could now see through sequencing.
What's next?
Ciro believes that the next step in the revolution could be could be: <de novo DNA synthesis>.
This technology could be the key to the one of the ultimate dream of biologists: cheap https://pubs.acs.org/doi/pdfplus/10.1021/acssynbio.6b00213[programmable biology] with push-button <species bootstrapping from DNA>[organism bootstrap]!
Just imagine this: at the comfort of your own garage, you take some <model organism> of interest, maybe start humble with <Escherichia coli>. Then you modify its DNA to your liking, and upload it to a 3D printer sized machine on your workbench, which automatically synthesizes the DNA, and injects into a bootstrapped cell.
You then make experiments to check if the modified cell achieves your desired new properties, e.g. production of some protein, and if not reiterate, just like a <software engineer>.
Of course, even if we were able to do the bootstrap, the <debugging> process then becomes key, as visibility is the key limitation of <biology>, maybe we need other cheap technologies to come in at that point.
This a place point we see the beauty of evolution the brightest: evolution does not require observability. But it also implies that if your changes to the organism make it less fit, then your mutation will also likely be lost. This has to be one of the considerations done when designing your organism.
Other cool topic include:
* <computational biology>: simulations of cell <metabolism>, <protein> and small molecule, including <computational protein folding> and <chemical reactions>. This is basically the simulation part of <omics>.
If we could only simulate those, we would basically "solve molecular biology". Just imagine, instead of experimenting for a hole year, the <2021 Nobel Prize in Physiology and Medicine> could have been won from a few hours on a <supercomputer> to determine which protein had the desired properties, using just <DNA sequencing> as a starting point!
* <microscopy>: crystallography, <cryoEM>
* analytical chemistry: mass spectroscopy, <single cell analysis> (Single-cell RNA sequencing)
It's weird, cells feel a lot like <Linux Kernel Module Cheat>[embedded systems]: small, complex, hard to observe, and profound.
Ciro is sad that by the time he dies, humanity won't have understood the <human brain>, maybe not even a measly <Escherichia coli>... Heck, even key molecular biology events are not yet fully understood, see e.g. https://en.wikipedia.org/wiki/Transcription_(biology)[transcription regulation].
One of the most exciting aspects of <molecular biology technologies> is their relatively low entry cost, compared for example to other areas such as <fusion energy> and <quantum computing>.