We are a Stanford student organization that runs out of the a teaching laboratory in the Shriram Center for Bioengineering and Chemical Engineering, and are dedicated to providing open space for anyone interested in pursuing bioengineering and biology inspired projects. You can see our website at biome.bio/ for information on what we do, or how to get involved with us!
I don't know what the odds are you will read this given the time difference, but I have a question that even my deepest of internet dives can't seem to answer. What if you don't want to use a pre-made backbone,? As in how are backbones originally made/designed? Is it just taking a blank plasmid and adding in promoters and resistance genes and the like as if they were lego bricks (in which case where does the blank space and unannotated resistance enzymes come from), it is a matter of extracting the very base from an actual yeast cell and sequencing and editing, or something else? Given the analogy of computer programming libraries, I'm curious to know what's actually happening in a library as I am using it. If you read this I was hoping you could point me in the right direction.
thank you for this very insightful video. a few questions: 1. crystallization or proteins process: a. how do you know that it is proteins and not something else that is crystallizing? b. how many proteins are in a crystal undergoing xray diffraction? if more than one, how can the computer analyze multiple proteins simultaneously (probably it isnt, but how do you know)? 2. xray diffraction: the crystal is being rotated. so the dots appear on the screen. how do you know which dots were produced at what degree of rotation? (it probably has something to do with Bragg's law?). is it important to know what degree of rotation each dot "represents"? other question: i saw another vid (British lab, which used Diamond complex for analysis) where they were preparing proteins for crystallization. there was a process called "dilution" of the proteins, to make sure there was a high concentration of protein. what is that all about? isnt a single crystallized protein necessary for the xray diffraction? (this of course is linked with another question stated above.) it is all so mysterious, that all of this is known, and that the technology of xray diffraction and computerized analysis has been developed with such a surety of the results! is it a kind of magic; that is, what are the larger implications about matter and energy here, and their relationship to mathematics? of course this is outside the range of your video, and perhaps your own interests. i wish you were my teacher!
i think the combination of the yellow and red is the average of the red and yellow individually in the first puzzle so that flourescence is at the mid way point of both no?
Hey Nauman, one way to think about biology is through physics models! Biophysics will model certain interactions like the binding of two proteases to a protease recognition site. They model based on outcomes. In our example there are three possible outcomes - either nothing cuts the site, red cuts, or yellow cuts. For example, if you have only red protease there are two possibilities: either nothing is cut or red cuts. The number of outcomes is limited to two! If both red and yellow proteases are there we can achieve all three outcomes. By a back-of-the-envelope calculation, Colin was thinking that 2/3 is greater than 1/2, so you'd expect more fluorescence because there is a greater chance of a cut outcome! But that's not what you see. Colin doesn't understand probability because if you have both red and yellow proteases the data shows that the likelihood that you cut decreases when you have both proteases compared to when you just have the yellow protease. We think this is due to the leucine zippers that bind the proteases together. This may render them less able to cut, therefore decreasing the fluorescent output. However, this is just our best guess. Let us know if you have more thoughts or questions! - Colin
Hi, I'm not sure if anyone's gonna be able to see this after two years, but why do you have all of those dead spaces with no annotations? I just see them everywhere on the plasmids I search for and I wonder if there's usually some purpose to all those dead spaces other than "missing annotations".
Hi - we did see it and great question! Having all of the components of the plasmid too close to each other may impact the binding of proteins to those sites so you want some space for easy access! So while the unannotated spaces do not transcribe to any RNA they are still useful and needed in a plasmid backbone!
the future here ;) we found the vaccine, but there occurred covid variations. We still have no freedom and will never get rid of covid :) THANKS you explain the theory better than my teachers of my university.
Evolution has to happen at the DNA level, not protein. A protein will not change shape or function without a different set of order form the DNA code. Herein lies the problem.
Thanks for a very detailed and informative session. I have a question or rather I would like to know your opinion on progress happening in the field of Proteinomics. On one hand we have companies such as Quantum-Si which are coming up with ways to detect the proteins by knowing the amino acids which are making up the proteins and on other hand, we have companies like Deepmind who are working on deep learning algorithms to guess the final shape of the proteins so that drug development can be accomplished. Based on your lecture, it is clear that both of the amino acid detection as well as detection of final shape have to go hand in hand. If my understanding correct ?
Your understanding is correct, they do go hand in hand! Given only the amino acid sequences of either existing or hypothetical proteins, Deepmind can predict the final structure with AI because there is a limited number of ways an amino acid sequence can physically configure itself, given the properties of each amino acid (ex. hydrophobicity) that act as constraints. This is one of the most, if not the most, powerful technology emerging in the field of proteomics right now, because what it means is very easy and fast design of proteins. Instead of going through an extended process of protein purification and x-ray crystallography to discover the structure of a protein, the AI allows researchers to have the likely final structure only after a sequencing step and the rest is done computationally. Knowing the structure of enzymes means researchers can rationally engineer them more effectively and rapidly. Additionally, when trying to construct new proteins/catalysts from scratch with only the final structure in mind, researchers can try different configurations and work BACKWARDS to find the amino acid sequence they want. All this purely done computationally. It will make engineering new proteins from scratch far more popular because before, there was no viable way to test how your brand new protein might fold and behave unless you put it on a plasmid, cultured it, isolated it, and ran some assays to test function and structure. So yeah, there is a lot of progress going on in the field recently! And it is moving very fast. Thanks a lot for your comment :)
