I'd like to add that we do generally have a pretty good understanding of where one gene "begins" and "ends", speaking of protein coding genes. Generally a protein coding gene is defined by what is known as an open reading frame(ORF), which is a stretch of nucleic acids that can be translated into protein, defined by the genetic code (the system of "rules" that govern the relationship between nucleic acid sequence and protein amino acid sequence). Nucleic acids are read in groups of three at a time, called "codons"(such as ATG, say), and a particular codon is translated into a particular amino acid. All these terms can be looked up on wikipedia. Genetic code, codon, protein biosynthesis. There are so-called "start" and "stop" codons that define the "beginning" and "end" of an open reading frame. Also, scientists have made synthetic novel genes with novel functions. There are examples of de novo designed proteins created by biochemists, that really do function, and were constructed essentially "bottom up". I'm surprised she didn't seem to know of this when she's working on minimal cells, but I supposed that's because she's a physicist and not a molecular biologist.

@paxdriver

2 жыл бұрын

She does know, but she also knows internally that there are still scientists who refute the results and there is likely more than one solution to those PDE's because biology isn't binary. Codons and telemeres aren't necessarily essential in a specific genome, it takes rigorous verification to snuff out all of the edge cases which hasn't been finished yet.

@Confuseddave

2 жыл бұрын

Genes and ORFs are not the same thing. In particular, while as far as I'm aware pretty much all vertebrate genes are ORFs (using a strict and narrow definition of "gene"), not all ORFs are genes, since there are many nonsense pseudogenes that fall under that category. Also, if you include organisms that have multiple polycistronic genes produced as a single mRNA transcript, simple the equivalence between an ORF and a gene breaks down further. The idea of an ORF is one useful term in grappling with what genetic information actually is, but it's important to remember that it's not a complete definition. I also thought she could have mentioned that when he asked a specific question about it, but the point she made in answer to his question was valid, and does not at all imply that she didn't know about ORFs.

I love when you branch out into other fields now and then 👍

Loved the talk. Sean im forever thankful for all the knowledge you bring to us

Another good one Sean, totally outside my usual subjects and very informative.

Thanks for another episode, Sean! Love'em.

As always, great podcast.

It took millions of years to develop the initial “minimal” cells from chemical building blocks. It would be interesting to explore the state of the science on how far we have come in actually understanding how proteins, base pairs and genes are made.

@rumraket38

2 жыл бұрын

We understand quite well how, inside living cells, proteins and DNA bases(and basepairs) are biosynthesized. These processes can even be made to happen, with the right molecules, outside the context of a cell, in test tubes(in vitro). You can look up nucleic acid biosynthesis(and RNA and DNA polymerization), and protein biosynthesis on wikipedia, the articles are pretty decent. We also have a pretty good understanding of how new protein coding genes evolve(so-called de novo gene evolution). For example, there's a type of mutation called a "gene duplication", which is literally what it sounds like. A piece of DNA is duplicated so you have the same thing twice. Once you've got two or more of the same gene, they can incrementally mutate and become more and more different from each other. That's one way to get new genes through evolution (there are many others).

Dualists, non-physicalists, hang on to your hats. I hope people refrain from making hasty statements like science will never figure out how life originated or science will never be able to explain consciousness and so on. Having said that sociologically we should keep ahead of science in terms of ethical use of science and make sure science remains accessible to all of humanity and extended living ecosystem.

Bumped into an industrial microbiologist the other day, and now this. Very cerendipitous and fun! Thanks!

Kudos to Dr. Strychalski!!! Just to throw this out, Stephen Wolfram's ideas about the nature of space maybe extremely well suited to describe cellular matrices (tissues, organs, etc.), and may help their engineering and production. The updating rule is the rule of biology which determine the cell cycle.

@paxdriver

2 жыл бұрын

She mentioned machine learning and alphafold, that's literally what ML does - matrices and scaler vector fields. The reason it takes so much research is doing ML irl means one experiment at a time, and repeating all of them. 100,000 features = orders of magnitude more physical experiments until the minimal cell/génome is solidified as known science. We're working on it though, it's a very good idea.

@nowhereman8374

2 жыл бұрын

@@paxdriver In my uneducated mind, when I looked at some of the example hyper graphs on Wolfram's web site seem very similar to some histology I have studied.

@paxdriver

2 жыл бұрын

@@nowhereman8374 sounds cool, I guess. Not sure I understand what you're saying tbh. Hyperspace geometry isn't useful as shapes, the pictures are just a visual tool to grasp higher dimensional relationships of data in math (ie cascading matrices aka tables within cells of tables). That's pretty much 90% of neural nets lol

@nowhereman8374

2 жыл бұрын

@@paxdriver I guess I have a little different take which is probably off in left field, but I will try to make this concise. In Wolfram Physics, I believe he is using the a potential, hypothetical hypergraph as the initial state of the universe. Space is defined by the relationship between unique points and is not defined by a traditional metric (e.g. a meter) but a rule. When the rule is imposed, the space changes which defines time... What I am saying is maybe Wolfram's approach is more suited to define tissues and organs. I believe the cell cycle defines the rule . www.wolframphysics.org/

@paxdriver

2 жыл бұрын

@@nowhereman8374 it's definitely interesting to ponder and I'm far from a specialist on the subject, but I think you'd be surprised to find were already employing layered fabric of data points in data sciences today as the standard practise of applying transformers, convolutional layers and the like. It sounds exactly the same as the shotgun approach with gradient decent applied to a cost function to steer or supervise the desired result. It's attaching and fixing edges to graphs in graph nets so it can even be built directly on geometric forms. You might enjoy 3b1b's series on neural networks if you haven't already. It's very well presented and doesn't skimp on the complexity while still being accessible.

great guest

Regarding lineage-agnostic, it kind of reminds me of the approach of Frances Arnold, who takes existing enzymes, tweaks them slightly in order to do perform a different function, and then places back it into bacteria in order to let evolution optimize and find a better solution to perform the same function. By breeding the best enzymes, she arrives with a highly efficient solution without needing to find the genetic-sequence for the complex-protein-folding that was required!

I guess where AI would be manipulating humans if AI want to build a body to live in.

Synthetic printed spherical cows when? Also, to simplify things, could we just skip a step and just print spherical hamburgers or would those not taste as good?

@gaurav63105

2 жыл бұрын

Already being done

31:27 Did she just call him Chris!??

@dominicklittle9828

2 жыл бұрын

That is very suspicious

Craig Venter not Ventner.

The sound quality is terrible.

Little substance here. I'm sure she is a brilliant scientist, but really didn't go in to much detail, but she talked around the subject rather than go in to much detail.

@paxdriver

2 жыл бұрын

Sounds like ought to start a podcast

Extremely interesting subject. Unfortunately I could not cope the whole way due to the bad sound quality. English is not my native language, so I might be whining here, but with nowadays technology available for everyone, I do feel a bit surprised. Anyhow, interesting subject.

@paxdriver

2 жыл бұрын

Most people don't have home studios. This quality should be the standard expectation, and anything better should be a bonus imho

@tommygrandefors9691

2 жыл бұрын

Hmm, I totally respect your opinion and answer, but I don't agree with you. I have frequent meetings with our teams in different countries around the world (group meetings). I don't use any headset and very seldom we encounter this type of audio issues. Most of the problems arise with teams in Africa due to glitches or temporary degradtion in bandwidth, but most often it gets solved by switching over to mobile network and re-joining our meeting from a phone. We simply don't experience this type of audio problems. No harm meant, but perhaps an overview of new existing technical communication platforms would do good.

@clayz1

2 жыл бұрын

@@tommygrandefors9691 I agree about the sound quality. To me it sounds like the result of digital compression to save bandwidth while transmitting. The audio is like the microphone is placed in the bottom of a tin can. It isn’t much, but it takes the edge off of what should be perfectly clear sound. I’ve noticed this on other podcasts. Interesting guest though.

@tommygrandefors9691

2 жыл бұрын

Yes, I totally agree with you. Since I don't use a headset or stand-alone microphone, I rely on the microphone in my laptop. Sometimes the software tries to be "too smart" when cancelling out noise. I get comments about this from my peers when I have a high load (RAM, CPU usage etcetera due to running too many VMs at the same time during the meetings for example). But in those cases my voice sounds far away, not like talking in a tin can. But I agree, probably more a side effect of today's technology (cheap microphones, talking in "wrong angle", low bandwidth, HW resource starvation). But perosnally I seldom experience this. Meaning that we should not accept this as "normal". 😉 Take care.

@user-qf3lq4zj8g

2 жыл бұрын

@@tommygrandefors9691 Sean actually mentioned a while ago that he used to send a bluemic to all invited guests (a HQ microphone). Perhaps this time was an exception or, as you wrote, the connection was very poor...

Didn't like the free mic then ?

The grey goo scenario is closing in I can only assume. I only wish I can have new set of teeth grown first.

I believe that childish laugh will become more of a mwahaha as she gets a little older!

Fif

It's 2021 and voice over IP is still bad :( I know there's usually a better mic but man

I hate to be that guy, but hey, FIRST!

@wayfa13

2 жыл бұрын

I too am very amped to watch/listen

@ccarson

2 жыл бұрын

Admit it, you loved it.

Not the sound quality for a podcast.

@paxdriver

2 жыл бұрын

It's free education with PhD leading researchers. Priorities dude lol

so, in laymans terms (!) you're poking about hoping for the best. hurry up, i'm reaching the end of my life, i would like to reavel in space before i die, so get these "make me younger drugs" going.

Sound quality is key to podcasts. But not this episode unfortuanately