The Systems Biology of COVID-19 and the SARS-CoV-2 virus. The class will build a foundation that includes the emergence of complexity, simple biological subsystems, their reductionist and equivalent toy and organ-chip models, and the measurements required to specify model architecture and parameters. Applications to biology, physiology, medicine, chemical and biological defense, pharmacology, drug discovery, and toxicology. UGrad: PHYS 240 01 and BME 290B; Grad: PHYS 326 and BME 395C.
Monday, September 3, 2018
SysBio18 Asgn_2A_Class_4_Article_02_2018_09_04
Sniffer-buzzer-1: Read, first pass: Article 02 J. J. Tyson, K. C. Chen, and B. Novak. Sniffers, buzzers, toggles and blinkers: dynamics of regulatory and signaling pathways in the cell. Curr.Opin.Cell Biol. 15 (2):221-231, 2003.. Post a PCRC on the Blog of your overview of this article.
Subscribe to:
Post Comments (Atom)
0. What I already know:
ReplyDeleteI have read this article in the past for the biophysics class I took this past spring semester. It covers the mathematical modeling of gene and protein concentrations. The article breaks bimolecular networks into repeatable units, much like circuitry. The mathematics are based on mass action kinetics. These models can predict the oscillatory nature of certain networks.
1. The most important thing I learned: Limit cycles are not all of the same species. There are "supercritical" and "subcritical" hopf bifurcations. In "supercritical" bifurcation the limit cycles do not change properties--they start at their stable operations which are of smaller amplitudes. On the other hand, "subcritical" bifurcation quickly builds up to larger amplitude, stable limit cycles from unstable ones. This has biological impacts in the quickness in which large amplitude oscillations occur.
2. My most pressing question: Through what manners do systems biologists accurately determine the constants necessary to build these network models. (Parameter evolutionary schemes? More interested in the biological techniques as I have no knowledge of these.)
3. Suggestion for class discussion: Building large biological functions from these base units (as in thinking through a simple example and proposing a model based on these units).
4. Thoughts/comments: The fact that sustained oscillations requires 3 factors is interesting. The end of the article mentions that communication between scientists is key to understanding bimolecular networks and systems. This begs the question as to whether the current scientific atmosphere is conducive to this.
0) OK
Delete1) OK
2) The determination of parameters is a giant problem. A lot can be estimated from biochemical measurements on reactions. We need to discuss this in more detail.
3) I think this will come in a bit.
4) I wonder whether there are oscillations within the scientific community...
0) What I already know:
ReplyDeleteI have had past experience in a lot of the biological examples that this article delves into with its differential equations. I have learned a lot about kinases, the cell cycle and its regulatory capacities, and many of these biochemical pathways. It was really interesting, however, to take a glimpse at these genetic and biochemical phenomena from the lens of a mathematician and physicist, with the article simplifying these proteins and such into the metaphor of electronics with signals and responses.
1) The most important thing I learned:
One of the most interesting and important things I learned about through this article is the concept of a sniffer and how it shows perfect adaptation to signal-response curves. For example, in Figure 1a the article discusses how a signal directly affects the response magnitude. However, if there is an external species X that parallels the signal in a second signaling pathway, the response magnitude changes. Even though there is a temporary response to a change in signal strength, the steady state protein concentration over time won't change in response to the signal strength. I really liked the article's example of the sense of smell, where temporarily you will notice a change in the smell in the room, but your signal response over time in steady-state adapts to it and you no longer react to the signal strength.
2) My most pressing question:
To what extent can signal-response curves and differential equations truly model a complicated genetic or biochemical network that involves many proteins, co-factors, inorganic elements, and more?
3) Suggestion for class presentation:
Utilizing molecular wiring diagrams to model evolution of proteins or organisms in an environment (given certain evolutionary pressures acting as the signals, while phenotypic change acts as the response)
4) Thoughts on class:
I am really fascinated by how the Tyson article set really examines biological phenomena like positive and negative feedback and frames it through a circuit like manner with a signal and a response. I am really excited to learn more in depth about each of the case scenarios presented in this article, and how differential equations and molecular wiring diagrams can model molecular biological phenomena.
0) Good!
Delete1) OK
2) Depends upon how complicated the process is.
3) We will look more at changing cell phenotype during development.
4) Good!
Kylie Balotin
ReplyDeleteReading Assignment 2a: Sniffers, buzzers, toggles, and blinkers
0) What I already knew
This article shows us models of biological systems with increasing complexity. I have seen some of these equations before in previous courses. For example, I’ve seen positive and negative feedback loops as well as Michaelis-Menten equations.
1) The most important think I learned
I learned about some new models though. I don’t think I’ve seen the perfect adaptation/sniffer equation. At least I haven’t seen it referred to with this terminology.
Sniffer: perfect adaptation and transient responses
Toggle: mutual inhibition by positive feedback and hysteric switches
Buzzer: sigmoidal signal response and switches
Blinker: negative feedback and oscillators
2) My most pressing question from the reading
Not so much a question as a comment, but I think the arrows in wiring diagrams in the figures could have been more clear. I wanted the dotted lines and solid lines to stand for certain types of reactions. For example, I wanted the solid lines to stand for synthesis. It took me a while to realize that the location that the arrow was pointing to was the important thing to pay attention to rather than the type of line that is drawn. This is even more confusing by the middle plot in the figures, where the dotted and solid lines do mean something. I don’t understand why they would use dotted lines and solid lines the way that they did in the paper.
How do researchers determine what the best model for a system is? Do they collect data and try to see which equations fit? What if they are not looking at the proper time scale and end up missing critical data?
3) A suggestion for a class discussion
We are going to be spending at least one class discussing this paper, but it might be interesting to have actual biologic examples of these models. The paper does a pretty good job at giving some examples, but maybe getting additional ones or going more in depth in the given examples might be interesting.
4) Any thoughts that you might have on the class or paper
I think this will be a useful reference for when we start modeling the gut-brain axis. There might be some useful equations that we can use or adapt for our purposes.
0) Good
Delete1) OK
2) I agree fully. As to time scales, yes, exactly!
3) Good point. Remind me to discuss this in class.
4) Yes!
0) What I already know: This is the third time I have seen this article in a course. I am also familiar with bifurcation theory and have heard about Turing patterns.
ReplyDelete1) The most important thing I learned:
I have read about it before but refreshing on the differences supercritcal Hopf bifurcations vs subcritical Hopf bifurcations vs. infinite period bifurcations was good.
2) My most pressing question:
The article touches on saddle node bifurcations, Hopf bifurcations, and infinite period bifurcations but how often do other bifurcations appear in biology? For example, do pitchfork bifurcations occur in biology?
The analysis done in this paper assumes the system is in steady state. Under what conditions is this assumption valid for biological systems?
What range does spatial patterning have in the gut/midbrain axis? Should we expect patterns just on the cellular level or will there be patterning at the tissue level?
3) Suggestion for class presentation:
Introduction to bifurcation theory for terminology and reading phase/bifurcation diagrams.
4) Thoughts on class or paper:
This paper should be approached carefully because it is difficult to understand phase/bifurcation diagrams without being exposed to it thoroughly. It is good that we are recreating the figures of this paper since it will help the class understand what the figures are actually plotting.
0) Excellent.
Delete1) Good.
2) I don't know of the examples. Some will appear in cell cycle. It might be worth looking for them.
I don't know of spatial patterning on the scale of the gut-brain axis. There may be, particularly in gut development, but, again, I don't know.
3) This will come out of the class presentations.
4) Yes!
0) What I already knew:
ReplyDeleteI am familiar with several of the biological pathways discusses in this paper, such as the cyclic AMP pathway.
1) The most important thing I learned:
An introduction to how biological systems can be modeled mathematically. All of the signal - response curves and various terminology such as bifurcation points and R-nullcline. I liked seeing how differential equations can be used to model biological phenomena.
2) My most pressing question:
The conclusion of this paper says that scientists need to learn to communicate to develop the sophisticated theory needed to understand molecular regulatory systems in cell physiology. How do we make this collaboration happen? Should scientists also learn things outside of their field to assist collaboration? But does this prevent scientists from becoming experts in their field?
3) Suggestion for class discussion:
Discuss how these models are discovered and developed. What techniques are used? I have no background knowledge on how people determined the differential equations for these signal-response models.
4) Thoughts:
This was a tough but fascinating paper for me. I love seeing the intersection of applied mathematics and physiology. I like seeing how traditionally separate fields can be meshed together.
0) OK
Delete1) Very good!
2) I think it is a matter of training. Classes like this address the problem.
3) We need to think about this later in the semester - KEEP ASKING THIS QUESTION!
4) Excellent
0: Knew: The activities of interacting proteins can be modeled using equations that dynamically approximate their abundances in relation to one another.
ReplyDelete1: Learned: I liked the use of graphs, and feel like this helped me understand/visualize the consequences of different types of modules/interactions and their scale. It's neat to see how simple interactions can be coupled in different ways to create such a variety of response curves.
2: Pressing ??: How do you fold in more complexity to this type of model in order to account for reactions that occur within different cellular compartments, net changes to the system that alter the efficiencies of various reactions, proteins which carry out multiple functions etc.?
3: Presentation: I would like to see where this type of modeling fits in to the ecosystem/spectrum of other techniques.
4: Thoughts: I'm guessing this falls within my definition of reductionism, I think it has the same power of just about any analogy in approximating what it describes within a narrow context -with the added benefit of being computer-friendly. When it comes to making models based on observations in real biological systems, I feel that major applications of contextualization or approximation will be necessary to help something like this work.
0) Good
Delete1) Excellent - an important thing to know and understand!
2) More terms, more compartments, more equations, more parameters to specify.
3) We will see that during the semester
4) Exactly!
What I already know:
ReplyDeleteMathematics of Dynamical Systems how to use bifurcation diagrams, nullclines and limit cycles to describe the dynamics of biological networks. I was also familiar with the different responses discussed from perfect adaptation, postive and negative feedback loops and their effects on the behavior of the system.
The most important think I learned:
That the basic components of signaling pathways and regulatory machinery can be modeled similar to circuits and that mathematical modeling can be leveraged to describe these behavior of these systems by assembling the many smaller connections together to describe the collective behavior. In a way that can be analyzed mathematically and experimentally.
My most pressing question:
Is there only one right way to draw the mathematical equations from the network represention or does the considering how the interactions in the network are mediated make more than one form of the system?
Suggestion for class discussion:
Is there any counter intuitive ways that two or more of these network motifs interact to get unique results that may not be apparent from modeling them seperately? (i.e. dynamical effects of cross-talk)
Thoughts on Paper:
Was a good introduction into building mathematical models of celular dynamics from at time-variant and time-invariant perspective. I just think it could of spent some time going into the assumptions they make in generating the forms of these different equations as that can impact the final resulting model. (ex: existance of protein complexes, principle of detailed balance kinetics, or entropic constraints)
0) Perfect - you can contribute a lot to the discussions!
Delete1) Good
2) I expect that in most systems that there are multiple different models possible. The trick is to identify the best, and that may be determined by how you plan to use it.
3) That will be interesting, but hard to bring about in class without a well-defined target. If you can come up with an example, then this will be useful.
4) These phenomena are beyond an introduction to the general concepts.
Knew: I was familiar with a good amount of the cell signaling pathways described here, but I had never thought about how useful it could be to describe them using a mathematical model.
ReplyDeleteLearned: Analyzing the pathways as circuits made the mathematical models much more clear to me- Most control systems that I've come across display very circuit-like behavior, so it makes sense that they can modeled so well by differential equations.
Question: How were all of these parameters specified? Is it based on empirical data? I would like to know more about how the models were generated.
Suggestion: We're already doing presentations on each figure and I'm really looking forward to learning from those.
Thoughts: Really cool paper! I feel like I need to read it a few more times to fully understand a few of the models but I love the intersection of physiology, electrical engineering, and mathematics.
0) This is GREAT!
Delete1) Excellent
2) Yes, empirical data.
3) Good
4) It definitely benefits from multiple readings!
0) Knew: I am familiar with the cellular processes mentioned in the paper (cAMP production, Cdk networks, etc.).
ReplyDelete1) Learned: The bottom-up approach to building complex network models out of elementary interactions such as "switches", "sniffers", and "buzzers" was helpful to break down the complexity of the overall model.
2) Pressing ?: Oscillatory systems that require three components are discussed, but I imagine there are higher-order oscillators that are more complex. would these just be combinations of the elementary interactions discussed in this paper? Or is this review not exhaustive?
3) Presentation: I'd like to learn about how these models were first developed/how parameters and constants are determined and agreed upon.
4) Thoughts: This paper provides a good way of breaking down signaling networks into simple loops and interactions, which allows you to piece together more complex models.
0) Good
Delete1) Good
2) There must be higher order oscillations!
3) You can go into the original articles. This has been an active area of research by many people for years. Tyson is in the lead.
4) Good.
4)
What I already know:
ReplyDeleteNot too much - perhaps just the notion of building complex networks from simpler units and the math tools used.
The most important thing I learned:
After making plots for some more ridiculous signal strengths, the basic mechanism of homeostasis still works for positive signal strengths! (it does break for negative S, though it would not make sense physically to have a negative signal? perhaps the lack of mRNA??)
My most pressing question:
The concept of a negative signal?
Suggestion for class discussion:
Just for kicks, could we somehow define a set of moods e.g. {M_1, M_2, ...M_n} and create feedback networks... perhaps mood can be also modeled in a similar fashion? E.g. if there's too much good news, then we get used to it and our mood is not heightened as much by it (sniffer?). How can we characterize mood variations throughout the day, the week etc over different time scales? Are there oscillatory cycles that we can analytically describe? (From life experience, there seems to be some...) It feels like someone ought to have already done something like this.
Thoughts on Paper:
John WikswoSeptember 11, 2018 at 4:34 AM
Delete0) OK
1) You can't have negative concentrations...
2) Negative rates. You can have an anti-signal, e.g., blocking antibody, but you can't take a concentration negative.
3) This would be interesting, but the problem is to obtain an objective measure of mood. It could be a qualitative tool, though!
For the article:
ReplyDelete0) What I already knew
Biology networks exist as combinations of genes, proteins and metabolites influencing each other.
1) The most important think I learned
Stable steady-state, unstable steady state, bifurcation points in mathematical meaning and their graphs.
Also in cell cycle (M-G1) negative feedback loop exists to break down mitotic cyclins.
Toggle switch – bifurcation point
2) My most pressing question from the reading
I have not quite understood the meaning of “buzzer” and its model. Also, during the M-G1 transition, what is the relationship between cdc20 and cdh1? How do they work together to promote cytokinesis?
3) A suggestion for a class discussion
Discuss the mathematical or engineering aspect of these models. Also the possibility of another toggle switch example in cell cycle.
4) Any thoughts that you might have on the class or paper
This paper is excellently written and provided me with overwhelming concepts of biological models. Even though personally more interested in the pathways it mentioned, it is good to have these tools in mind.
0) KNEW
ReplyDeleteThis was my second time seeing this article. I knew the general forms of most of these types of models. I also knew the basics of wiring diagrams and a bit about protein kinetics.
1) LEARNED
Steady state patterns (Turing patterns) form from spatial signaling in a time-independent manner. I also learned (and finally understood) Michaelis-Menten kinetics
2) QUESTION
I am still a bit unclear on the full functioning of toggle switch, so I need to see that presented in class.
3) DISCUSSION
4) THOUGHTS
I like the Tyson article. Obviously very in-depth. This is the first time that I've gotten a full understanding of the interplay between the middle and right graphs of figures 1 and 2.