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.
0. Knew:
ReplyDeleteI knew how GRN determine the epigenetic landscape, and how attractors in this space represent stable cell fates. I knew that the dynamics of the GRN is determined by how the gene expression levels change over time, not the actual network itself changing (basically everything from the previous two Huang papers).
1. Learned:
I learned how Neo-Darwinism, the linking of natural selection and Mendelian genetics, is challenged by epigenetics and GRNs in the sense that genotype is not always a direct correlation to phenotype. Instead, many traits of organisms, for example the fractal tissue pattern in the lung, can be explained through chemistry, physics, and mathematics instead of natural selection.
Even though from the previous papers I understood that the GRN does not change for a given organism, I gained more understanding of this idea. The paper presents two time scales: the network dynamics that take place during an organism’s lifetime, and a rewiring of the network that arises from mutations and is on the evolutionary timescale. Therefore, network dynamics refers only to a change in gene expression, while an actual change in the network does not occur for a single organism; rather, it occurs through mutations through “generations” of organisms.
Other than that the paper largely coincided with the previous paper. The additional information obtained from this paper was how GRNs modify Neo-Darwinism to explain why we do not see a 1:1 correspondence from genotype to phenotype.
2. Pressing Questions:
The paper discusses how perturbations can move gene expressions to a different attractor state, but the perturbations do not change the landscape itself, but could you, in fact, describe the perturbation as a temporary change in the landscape (much like providing an additional force would change the potential energy landscape)? Would this even be useful? Part of me feels like it would not, based on the discussion of biology group B in box 1, but I have been wondering this.
Also, the paper describes perturbations as being caused by gene expression noise. Are there any other types of perturbations that can occur for the GRN?
3. Presentation Topic:
Going along with my last question, examples of other perturbations, if there are any, would be an interesting and helpful presentation.
4. Thoughts:
This topic is my favorite topic so far. This paper does a very good job at laying a base understanding of GRNs’ topology and dynamics and how these relate to phenotypes.
0.Knew: I was already familiar with Darwin’s theory of natural selection, the Mendelian concept of discreet genes as the vehicles of inheritance and the linear causation of one gene to one trait. I knew about the central dogma.
ReplyDelete1.Learned: I learned that epigenetics is the idea of non-genetic contributions to determination and inheritance of traits. I learned that scientists named Gould, Lewis and Goodwin contributed to modern scientists being aware of the fallacy in natural selection being the only influence on phenotypic traits. I learned more about how epigenetic landscapes were constructed. I learned that the difference between network structure and network dynamics is that the former is a static collection of interconnected genes, and the latter refers to the collective behavior of these genes that changes expression level with time. Returning to the historical aspect of the paper, I learned that scientists named Monod and Jacob discovered that genes can be regulated by the expression of other genes. It was also interesting that the epigenetic phenomena of genetic expression noise and the source of variability that is the GRN directly challenge Darwinian evolution. I learned that not all gene expression patterns exhibit equal stability and that a network’s stable state is called an attractor state because it attracts surrounding states that are not as stable. I learned about the effect a mutation can have on the physical epigenetic landscape.
2.Pressing Questions: If genes are not the sole basis of inheritance, what else can possibly store our directions for proteins synthesis? (p.149)
What are some examples of gene expression patterns that are more and less stable? How long is the life of an unstable gene expression pattern? Does it depend on the degree of instability? (p.154)
3.Presentation: I think I need a refresher on epistasis. Maybe a short presentation on this would be worthwhile.
4.Thoughts: This paper was comfy to read because I am more familiar with genes and gene expression than some of the other topics we have covered, and it was certainly easier to understand this paper compared to those that are saturated in math. I also think the epigenetic landscape picture was pretty useful, and I appreciate that this article provided historical context.
I agree that epistasis would probably be a nice thing to go over. I thought it was really provocative that he said epistasis was "phenomenological" on the last line of the first page. Does he really mean to say that it's not a real thing, and that biologists have been thinking about it wrong this whole time? Is it some sort of band-aid on the Neo-Darwinistic paradigm?
DeleteKelly McGee
ReplyDeleteArticle 12a
The molecular and mathematical basis
of Waddington’s epigenetic landscape:
A framework for post-Darwinian biology?
0: Knew: Having read the papers for last Thursday and having also read 12b already, I was very familiar with the idea of quasi-potential landscapes and attractor states, and their applicability to gene regulation networks. I was also aware of the historical formation and general concepts of neo-Darwinism.
1: Learned: I learned about gene regulation networks' applicability to neo-Darwinism as an important expansion to biological theories of evolution from a high-level view. I learned about a number of semantic arguments in systems biology including the word "network."
2: Questions: p.154, first column, last paragraph: How did the word "zillions" make it into a peer-reviewed publication?
Also on p154, I think that in discussions regarding the genome's ability or inability to select a given gene expression profile, we must also consider, in some way or another, what the sperm and egg cells contribute in terms of creating the "ground zero" of a particular genomic instance's gene expression. Are there any studies that address this?
3: Presentation: If the math and physics-heavy people would think it helpful, a brief review of neo-Darwinism.
4: Comments: While this paper was thoroughly interesting, and could address potential applications for homunculi and SOC's to be used in evolutionary biology studies, I question this particular paper's applicability to our general course of study. Our class' view seems to be focused on the current and immediate (0-10 years ahead) state of systems biology, and I don't see any particular usefulness of this paper within that view of study.
I also wonder how important a solid understanding of the epigenetic landscape is to studying mechanisms of action. But I guess it's important to conceptualize how things change in the body before we try to understand the complicated changes a drug or something might introduce.
DeleteAlternatively (here I selfishly plug my own idea), maybe we can use analogues of these ideas to understand cell death in MOA stuff. That is, maybe thinking about a 'cell death landscape' will be helpful.
Class 12. Article 13. S. Huang. The molecular and mathematical basis of Waddington's epigenetic landscape: A framework for post-Darwinian biology?
ReplyDelete0. KNEW:
This is the third paper we've read from Sui Huang on this topic, and to be honest it seems like we've seen all of this already. Does he write one of these every couple of years? This one is dated later than the previous two, but seems like it has about the same amount of content.
1. LEARNED:
I guess the most interesting thing here is pointing out that gene regulatory networks are not rigidly compatible with what Huang calls the 'Neo-Darwinian' paradigm: mostly 1-1 gene expression-phenotype, in accordance with the so-called central dogma of molecular biology (DNA -> mRNA -> protein). In other words, the argument is that biologists tend to think about the genotype-phenotype relationship naively, ignoring a lot of the crucial epigenetic machinery at work under the surface.
Really, a lot of this was a retread of what was discussed in the previous qualitative exposition-type paper (the epigenetic landscape can be thought of as a ball rolling around in a rugged landscape, this gives you a way to think about stability, networks are better thought of as a time-invariant background on which gene expression levels change...). This is a little disappointing considering the word 'mathematical' in the title made me hope there was some more explicit detail here on the dynamical systems stuff he keeps talking about.
2. QUESTIONS:
The other paper answered my qustions about the implementation of and possible usefulness of this framework. However, looking at Box 1 (pg. 152) made me think of a tangential question I find kind of interesting.
Huang makes a big fuss about treating networks as time-invariant, and gene expression levels as time-dependent, as being the 'right' way to think about things. He thinks it is unfortunate that there is a group of biologists which instead consider the network time-dependent.
In quantum mechanics, there is an analogous situation: one is interested in abstract spaces of functions, and operators that act on these spaces. You can either choose the coordinate axes to be time-independent and the operators to be time-dependent (the so-called Schrodinger picture), or you let certain operators be time-independent, and think about the coordinate axes as moving around with time.
Actually, now that I think about it, there is an analogous idea in classical mechanics too. Anyway, both points of view (axes TI, operators TD; axes TD, operators TI) are equally valid, and there is a well-defined mathematical way to transition between them.
I wonder if in this case there is also some sense that both points of view are equally valid. In that case, as with QM, I further wonder if there is a clear way to transition between them, and if considering each one might be useful in different circumstances.
I could be pushing the analogy too far. But it's interesting to think about!
3. PRESENTATION:
Some words on 'Neo-Darwinism' and the possible 'post-modern synthesis' Huang speaks about are probably all that is really necessary here. It's good terminology to have in framing why people care about gene regulatory networks and systems biology; people care about them because they're important and haven't traditionally been thought of as important!
4. THOUGHTS:
I am trying not to talk trash, since I thought the last two Sui Huang papers were nice, but this one is probably one too many on the same exact topic.
Chinowsky_TheorSysBio_PCRC_100416
ReplyDelete0. Knew
From the papers we had read previously, I am familiar with the concept of quasi potential wells and attractor states as model of the “stability” of a GRN. The previous discussions on network structure and its basis in the entire genome (regardless of what is actually expressed) were also helpful for understanding this paper.
1. Learned
Learned that “neo-Darwinism” based biology requires a framework that is more complex and dynamic that that of the old paradigm of genetics. Our deeper understanding of the genome, epigenetics, and the molecular basis of gene-gene interaction. Throughout the paper, there was an emphasis that genotype does not directly lead to phenotype, something that is very much relevant, and has only become more true as our understanding of genetics continues to increase.
2. Pressing
What exactly constitutes “epigenetic phenomena”(pg 150)? Also, for a paper with a title that implies an exploration of the mathematical basis, there seems to be very little hard math in the body of the paper itself.
3. Presentation
The “evolution” of genetics and GRNs
4. Thoughts
It didn’t seem to me like this article presented many new concepts compared to the readings we have done in the past few weeks.
0. KNEW:
ReplyDeleteI was familiar with many of the concepts and theories that were presented in the introduction. I feel comfortable with the theory of natural selection as well as Mendelian inheritance, and thus have a decent grasp on the neo-Darwinism principle. I was also familiar with the question that is posed against this principle from other readings that mentioned epigenetics. I knew about attractors and epigenetic landscapes.
1. LEARNED:
I learned about the GRN and the dichotomy associated with the terminology. The different descriptions of networks helps clarify the thoughts of the author, but I am confused why the fixed terminology is preferred. I added to my knowledge of the quasi-potential landscape in terms of a potential energy well.
2. PRESSING QUESTIONS:
Is it possible to build a master knowledge in a scientific subject without first using constraints? Along those same lines, is systems biology a means to an end for the collection of information on the broader subject, or is it only another stepping stone in the evolution of biology?
3. PRESENTATION TOPIC:
Waddington's background and contributions.
4. THOUGHTS:
Many similar ideas. I would have appreciated figures that presented concepts with a different perspective than the previous papers.
Ben Terrones
ReplyDeleteSysBio16 Asgn_12A_Class_12_Article_13_2016_10_04
0. KNEW:
I knew that there are many issues with assuming a linear, causal relationship between genotype and phenotype. I also knew about most of the historical view of biology and Neo-Darwinism, and the recent focus on GRN’s and the “epigenetic landscape” and what these terms mean. I knew that a specific phenotypic state is described by the gene expression pattern at that time and there are constraints on the realization of expression patterns from inter-dependencies of gene expression. I also knew about attractor states from previous papers.
1. LEARNED:
The clarification about how the “wiring” of the gene network doesn’t change, just the expression levels was very helpful in my understanding. I also learned how the perturbations can change the trajectory if large enough and how the cells “remember” the expression profile of the new attractor state. Mutations change the network architecture, non-mutagenic perturbations change the expression profile.
2. PRESSING ?:
What are the Baldwin effect and Neo-Lamarckism that were mentioned in the conclusion?
3. PRESENTATION:
A presentation on the epigenetic phenomena presented by critics of Neo-Darwinism that were mentioned in the conclusion: (Waddington’s genetic assimilation, Baldwin effect, and Neo-Lamarckism).
4. THOUGHTS:
I enjoyed this paper because it was a nice review of what we have been talking about the past few classes. The author explained things well and clarified some things I was confused about.
0. Knew:
ReplyDeleteI thought this paper recapitulated what was read and discussed in last Thursday's class. I came away from that class with a pretty good understanding of state space and how to interpret the topological landscape.
1. Learned:
I hadn't considered the relation of epigenetics as we've discussed in light of Darwinism. This paper also drove me to do a little refreshing on the concept. This paper did a good job describing the constraints and cleared that up for me a little more and the paper also clarified for me how each cell has the same GRN.
2. Pressing ?:
Pg. 156 I was confused about how some phenotypes are directly encoded by a mutated gene (and the paper references haemoglobin and melanin synthesis). What exactly does this mean? Probably a basic biology question.
3. Presentation:
Attractor state memory.. explain how this emerges and some examples
4. Thoughts: Good paper to recap the previous and give another relation where using this thinking could be applied. I was surprised the paper only contributed 1 short paragraph to explaining the mathematical description of state space.. I guess it's assuming the reader has some background and/or can really interpret the details from the figure.
0. Knew:
ReplyDeleteI thought this paper recapitulated what was read and discussed in last Thursday's class. I came away from that class with a pretty good understanding of state space and how to interpret the topological landscape.
1. Learned:
I hadn't considered the relation of epigenetics as we've discussed in light of Darwinism. This paper also drove me to do a little refreshing on the concept. This paper did a good job describing the constraints and cleared that up for me a little more and the paper also clarified for me how each cell has the same GRN.
2. Pressing ?:
Pg. 156 I was confused about how some phenotypes are directly encoded by a mutated gene (and the paper references haemoglobin and melanin synthesis). What exactly does this mean? Probably a basic biology question.
3. Presentation:
Attractor state memory.. explain how this emerges and some examples
4. Thoughts: Good paper to recap the previous and give another relation where using this thinking could be applied. I was surprised the paper only contributed 1 short paragraph to explaining the mathematical description of state space.. I guess it's assuming the reader has some background and/or can really interpret the details from the figure.
Natalie Hawken
ReplyDeleteAsgn 12A, Article 13: The molecular and mathematical basis of Waddington's epigenetic landscape
0. KNEW
From the previous articles, I knew about the GRN and the assumed permanence of the network. Also, the other Huang articles discussed the "epigenetic landscape" and how the cells will fall into the basins of attractions as they go down to the lowest quasi-potential. Also, I knew how the noise in the gene expression explained the heterogeneity of cell populations while still remaining in the same basin of attraction.
1. LEARNED
In this article, the ideas about the epigenetic landscape and the GRN are applied to new theories of natural selection and evolution that differ from the Neo-Darwinism. I wasn't aware that Neo-Darwinism was starting to be replaced by "Post-Modern Synthesis" which discusses how complex organisms are not just based on genes, but also are affected by laws of physics, physical constraints, and self-organization. This article also helped to clarify the constraints of gene-gene relations greatly limit the combinatorics of gene expressions, which minimize the possible number of distinct cellular phenotypes.
2. MOST PRESSING QUESTIONS
On pg. 156, the author states that the quasi-potential landscape is a "mathematical entity" but doesn't clarify what this entity would be. What are some examples of equations used to represent the landscape?
Since the author states that genes alone cannot account for all inheritance, are there any well-regarded twin studies that accurately map what other outside sources could lead to variability in inheritance?
3. PRESENTATION TOPICS
The paper mentions in Fig. 1 "biocomplexity." Have a presentation about the development of this idea/field and the major contributions it made to the Post-Modern Synthesis view of evolutionary biology.
On pg 155, the author notes how attractor states "convey a memory" of gene expression for a specific state. How does this memory work? How long does it last? How accurate is it in recalling the correct attractor state?
4. THOUGHTS
This paper was good for a general overview of the two papers from Thursday, but I found it kind of boring because very little new information was presented. Also, most of the claims were theoretical and didn't include quantitative analysis. I would have preferred to read an article about attractor states or the epigenetic landscape if it focused on a specific cell type or lineage and actually included wet-lab data to support its claims.
0. KNEW
ReplyDeleteI was familiar with the guiding principle of Neo-Darwinism, namely “gene —> mRNA —> protein (=trait)—after writing that mantra for an extra point on every biology test in high school, it’s become carved in stone in my memory. I was also familiar with the fields of epigenetic (modeling gene interactions, as we have been talking about extensively lately) as well as
1.LEARNED
The first striking point of the paper was the phrasing that genes and mutations of such are not the sole basis of inheritance. Rather, natural selection can also be influence by the two epigenetic phenomena of a multitude of GRN states as well as gene expression noise.
The paper makes a point of defining in plain terms the discrepancy between network structure and network dynamics (not to be confused with dynamic networks). Network structure refers to a hardwired, time-invariant architecture that is “carved in stone” and does not change during a lifetime for a given genome. Network dynamics refers to the change in expression levels of genes that take place in an organisms lifetime without changing the network structure. The only change of the structure that takes place occurs on the evolutionary timescale.
The paper then delves into gene-gene regulatory interactions acting as constraints and the relative stability of each state in which certain genes are expressed. The genome favors certain states to others in order to minimize quasi-potential energy U for the state. Stable states are called attractor states that attract less stable states to its basin of attraction (local min for U). The landscape represented by N dimensions for N possibly expressed genes plotted against a U axis does not change, and shows basins and hills where favorable and unfavorable network states lie, respectively.
2. PRESSING ?
What does a single genome refer to— one organisms genetic information? Can there be done any modeling of genetic mutation with any regularity as can be done with gene expression?
3. PRESENTATION
gene expression noise
4. THOUGHTS
I thought this was a great paper. It presented difficult conceptual material in a visual, comprehensible way.
0. KNEW
ReplyDeleteI was aware of epigenetics as an improvement over the linear theory of genotype-phenotype. The epigenetic landscape figure was familiar to me because of the prior Sui Huang papers that we have read.
1. LEARNED
There was a lot of clarification on the interpretation of epiegenetics thanks to the supplementary in the back. I have a good understanding of GREs now and how differences in expression levels rather than the structure of expression itself that determines the epigenetics of an organism.
Networks and noise, in turn, are the non-genetic sources of diversity and stochasticity that are the drivers of Darwinian evolution.
2. QUESTIONS
Why isn't there as much math as the title may suggest?
3. PRESENTATION
Clarification on Neo-Darwinism, Lamarckism, etc.
4. THOUGHTS
This was a short and sweet paper that clarified several concepts from the other Sui Huang papers.
Sylvia Morrow
ReplyDeleteAsgn12A_S. Huang. The molecular and mathematical basis of Waddington's epigenetic landscape: A framework for post-Darwinian biology?
0. KNEW: The majority of this article was very familiar because of A11 and A12: gene regulatory networks, epigenetic landscape, attractors.
1. LEARNED: Because this article seems to have been written in part to clear up the discourse regarding terminology, I found it very helpful that Huang went through many things very explicitly (what they meant, what they implied, what they didn't mean); for example, the linear causation scheme, the relationship between natural selection and epigenetics, gene regulatory network (in particular, wrt dynamics), and developmental constraints.
2. PRESSING ?:
--on pg 153 Huang says the GRN dynamics and the gene expression noise are non-genetic sources of diversity and stochasticity. by "non-genetic" does he just mean non-monogenetic? or am I missing something here?
--only attractor states are considered associated with phenotypes. is this because the system won't stay in a non-attractor state long enough for it to show up as observable characteristics? or do phenotypes by definition need to be stable?
--not super relevant but I am curious: Huang talks about mutations as being a "spurt of evolutionary progress". how often do we see mutations that are harmful? are some diseases caused by specific mutations? is that what makes some diseases hereditary?
3. PRESENTATION:
--A specific example of the genes A,B,C example on pg 154 in the section "Network dynamics reflects constraints on the realization of expression patterns"
4. THOUGHTS: Was interesting to read these three papers by Huang all on the same topic. His argument a little more fleshed out in each one.
Knew 0
ReplyDeleteI was familiar with the concept of GRNs from the papers we've discussed thus far in class. The concept of epigenetics was also familiar from my biology classes, though it has recently expanded to include the broader definition outlined by Huang last week.
1 Learned
This paper helped draw the distinction between the network structure and the network dynamics and the relevant time scales on which they operate.
2 Pressing Questions
If the structure of the GRN can be mapped onto a epigenetic landscape, can epigenetic landscapes be mapped back onto a unique GRN? IE, can observed patterns of gene expression be used to determine the underlying structure of the network?
If GRNs are so constrained, does that allow for use to simplify our network specifications and allow for increased specificity in our models (more degrees of freedom). How can we rigorously simplify our network representations (eigen genes, or is there a better way?)
3. Presentation
Maybe a discussion of specific examples of evolution of GRNs
4. Interesting paper, would have provided good context for last week's readings.
Stephen Lee
ReplyDeleteClass 12, Assignment 12a
The molecular and mathematical basis of Waddington's epigenetic landscape: A framework for post-Darwinian biology?
(0) Knew:
I approached this paper with the knowledge I’ve accumulated from the previous few lectures (mainly Zach’s earlier presentation on the epigenetic landscape and attractor states. I have a fairly substantial knowledge of evolutionary theory.
(1) Learned:
I learned that epigenetic phenomena are substrates upon which natural selection acts and may accelerate evolutionary change. I also learned that a Post-Darwinian framework for evolution will be informed by mathematical models that justify different network states based on a quasi-potential quantity.
(2) Pressing Questions:
Could this paper make a compelling argument for massively parallel assembly of GRNs? It seems to me that the authors are somewhat skeptical on this front. Can such a high degree of stochasticity be boiled down into an empirical model?
(3) Presentation Topic:
Reconciling stochastic effects with high-throughput GRN assembly
(4) Thoughts:
This was an interesting paper. Looking forward to presenting on it!
0. KNEW:
ReplyDeleteI knew about the concept of natural selection, how genes give rise to phenotypes, and Waddington's idea of the epigenetic landscape.
1. LEARNED:
I learned about the concept of a GRN and how this is disruptive to genotype-phenotype mapping. I learned that the network should be thought of in concrete terms of possible relationships, not transient states. I learned about the difference between network structure and dynamics.
2. PRESSING ?:
How much does epigenetics actually account for variations in gene expression?
3. PRESENTATION:
I would make sure to graphically shown the relationship between GRNs and epigenetic landscapes.
4. THOUGHTS:
This was a good paper for providing a background for the various types of networks and their relationships.
0 Knew
ReplyDeleteI was familiar with GRN from last week's papers. I was also familiar with epigenetics in the context of cancer biology.
1 Learned
I appreciated the explanation on GRN phylogenetics. ("There is ONE, and only ONE, static network topology (=structure=wiring diagram) in a given genome.") It clarified some of the confusion I had on the semantics of epigenetics last week.
2 Pressing
How do you model these reversible attractor states? Using stochastic differential equations? I am also curious about the statistical thresholding used to define phenotype. Are there models that utilize gradations rather that Boolean variables?
3 Presentation
Attractor state memory
Phenotype statistical thresholding
4 Thoughts
Interesting paper, look forward to seeing how class discussion develops since it builds on last week's reading.