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.
Chinowsky_TheorSysBio_PCRC_113016
ReplyDelete0. Knew
I have some basic knowledge of genetic interactions and genomic sequences, as well as network mapping, thanks to previous class discussions.
1. Learned
This paper defines a positive genetic interaction as a double mutant with fitness greater than expected, and a negative genetic interaction as a double mutant with a fitness that is below what is expected. The mapping of genetic interactions was done using a genome scale library of yeast genes, and both essential and non-essential genes were examined. Individual genes can have a specific genetic interaction profile, which is the set of negative and positive interactions. As one might expect, genes involved in similar biological processes, and genes encoding proteins that function in the same pathway, show similar genetic interaction profiles. Genetic interaction profiles can be used to look at functional hierarchies as well.
2. Pressing
What is an interaction that is neither positive or negative? Where the fitness is as expected?
3. Presentation
Genetic interactions within protein complexes, at a more specific level that what this paper deals with
4. Thoughts
I feel like I got lost while reading this paper. There was so much information, I don’t think it would be possible to absorb in one read.
Yes, this is a dense paper, but it is important to learn how to separate the specific details from the larger concepts, many of which reflect discussions in yesterday's class and before. It may take much more than one reading to understand all the details, but we don't need to understand all the details. In fact, there is merit to ignoring the details so as to see the larger picture. Hence given the time, one reading may have to suffice. Hence, I read the summary a couple of times, worked to understand the key figures, and then started identifying key themes. In many ways, this paper reminds me of Rachmaninov's Rhapsody on a theme of Paganini -- 24 variations of a catchy tune that I heard last weekend. So the trick is to pick out the basic tune.... I think I now have a general understanding of how this paper relates to our earlier discussions! We can go over what each of you pull from the paper, but I believe that there is a clear theme when viewed from the context of our entire class! Let's see who can find it!
ReplyDeleteBen Terrones
ReplyDeleteSysBio16 Asgn_25_Class_26_Article_22_2016_12_01
0. KNEW:
I knew from previous classes about different methods of network construction.
1. LEARNED:
This paper focused on genetic interactions, positive interactions being those that produce a phenotype better than expected, and negative interactions those that cause unexpected cell death. The negative interactions are of particular interest because of the potential for identifying therapeutic targets. They identified about 550,000 negative and 350,000 positive gene interactions in their experiments. These were used for the primary goal of the paper, which was construction and analysis of a genetic interaction network using essential gene pairs from yeast cells. Essential genes are densely connected hubs that form the scaffold of the network. They were also able to sort out a functional hierarchy based on similarities in genetic interaction profile. They concluded that using the global genetic network, it might be possible to uncover genotype-to-phenotype relationships.
2. PRESSING ?:
It’s not clear to me how the negative genetic interactions would be used to identify antibiotic or cancer therapeutic targets? What is CRISPR-cas9 genome editing? In the discussion section it said that CRISPR-cas9 genome editing approaches offer the potential to map global interaction networks in humans, but has it been done? How much harder would it be?
3. PRESENTATION:
Mapping global interaction networks in human cells
4. THOUGHTS:
I really enjoyed this paper; it made me think. A lot to take in, but the figures were neat. Cute as Dr. Wikswo would say.
0. Knew:
ReplyDeleteI knew about gene regulatory networks. I knew about using the Pearson correlation coefficient as a way to quantify similarity between gene expression levels and interactions. I knew that gene interaction networks could be used to describe and predict functionality of groups of genes.
1. Learned:
This paper described constructing genetic interaction networks for nonessential, essential, and global genes found in yeast. It is found that based on how these genes cluster through various analysis techniques, the modules formed are able to describe and predict different cellular functions, phenotypes, and protein complexes. These networks were also able to identify the roles of uncharacterized genes. Considering both positive and negative interactions was essential to the analysis because the information obtained from each type of interaction identified different information. The analysis and conclusions drawn could be very useful in studying the gene interactions of other organisms.
I learned that a pleiotropic gene is a gene that influences two seemingly unrelated phenotypes. Pleiotropy in this paper was quantified as genes with a high degree of negative interactions with a score that measured the functional breadth of genetic interaction that profiles associated with the pleiotropic genes.
2. Pressing Questions:
Has this type of network construction been done with other organisms? I know the discussion section mentions that the conclusions drawn here should be conserved in humans due to previous observations, but shouldn’t this still be fully looked at (or is this a current thing)?
Are the genetic interaction networks described in this paper the same thing as the gene regulatory networks we have previously discussed, or are there differences?
3. Presentation Topic:
Examples/evidence of genetic conservation between yeast and humans.
4. Thoughts:
This was a great paper to finish the class with. I was able to see how much I have learned because I knew before this class I would have had a hard time getting through this paper. Learning about different hierarchical clustering methods and gene regulatory/interaction networks was helpful in understanding this paper.
Global Genetic Interaction Network (Costanza et. al.)
ReplyDelete12/1/16
0 (Knew): We had already covered the development of hierarchies based on level of similarity, which is a major aspect of this study.
1 (Learned): I learned what positive and negative mutation interactions were, and how analysis of these interactions can allow for fairly straightforward development/addition to gene interaction networks, as well as the creation of interaction hierarchies. Additionally, I learned that loose similarity in such network is tied to general location (compartment) in the cell, while more specific similarity relates to specific functionality.
2 (Questions): It says this is in the supplementary materials, but I'm still a little confused as to how they conducted the initial study that gave them the data they spend most of the paper analyzing.
3 (Presentation Suggestion): As Aaron touched on, a rundown of a similar study in humans, or, if one does not exist, an explanation as to the difficulties in running such a study on human cell lines.
4 (Comments): Figures 2 and 3 seem the most relevant to our class, much of the second half of the paper (dealing with positive/negative interactions.
0. KNEW:
ReplyDeleteI had previously learned about network construction and the purpose of mapping these specific interactions. I knew how nodes and edges are involved with specific algorithms in constructing these networks. I have heard of CRISPR editing (I believe there is a poster on it directly outside our classroom).
1. LEARNED:
Through Figure 1 I learned that genetic information was being mapped with nodes being pairs of genes and edges being interaction profiles. The three maps in this Figure (A,B,C) all show different networks of genetic information that involve essential and nonessential yeast genes. Essential genes are "network hubs" that display five times as many interactions as nonessential genes. From 5416 genes, 1 million interactions we identified (550,000 negative, 350,000 positive). A positive genetic interaction describes a double mutant exhibition that is greater than expected based on the combination of two single mutants, while a negative interaction describes a double mutant defect that is more extreme than expected.
2. PRESSING QUESTIONS:
In the discussion it is mentioned that the genetic network was mapped under a specific conditions, and if these conditions are changed, new interactions will be revealed. I am confused as to what these "two key factors" are.
3. PRESENTATION:
An overview of these figures and the paper in general as the text was very dense.
4. THOUGHTS:
Very dense, but I can't imagine how little I would have gotten out of it if we hadn't spent so much time on the other papers throughout the semester.
0.Knew: I knew that mutations in two or more genes could create an unexpected phenotype. I knew the importance of using networks to better predict gene and pathway functions based on how closely they are correlated with known genes and known pathways.
ReplyDelete1.Learned: I learned the difference between what was considered to be an essential and a nonessential gene in this study. I learned that synthetic genetic array (SGA) analysis automates the combining of defined mutated genes and helps quantify analyses of genetic interactions. I learned that synthetic lethal interactions can be used to identify new antibiotic/cancer therapeutic targets. I learned about the difference between a negative and a positive genetic interaction. I learned that nonessential gene interactions are indicative of mitochondrial function or vacuolar transport, and essential gene interactions are indicative of chromosome segregation and RNA splicing. I learned that SAFE identifies dense network regions that are associated with certain functions. I learned about phenotypic capacitors.
2.Pressing Questions: The article mentions near the beginning that genetic interactions “may also explain a considerable component of the undiscovered genetics associated with human diseases.” Which diseases in particular that remain to be explained in such a way?
Also, like Aaron, I am inquisitive as to the progress of this type of networking for other organisms.
3.Presentation: A (comprehensive) review of the figures involved would be very helpful.
4.Thoughts: This article was pretty dense for me. I am hoping that the class discussion can help me parse through the information and make it more understandable.
Asgn 25, Article 22: A global genetic interaction network maps a wiring diagram of cellular function
ReplyDelete0. KNEW
I knew about loss-of-function studies in terms of knocking out genes and then determining if there was a subsequent loss of viability or decrease in functioning. I knew that there were specific genes/proteins known to be the central effector in specific cellular pathways and functions and genes that acting in multiple pathways (pleiotropic). And I knew that there is a lot of redundancy in the eukaryotic genome in order to prevent total loss of function if one protein is mutated. We learned about Pearson correlation coefficients in class, but my understanding is still a bit murky.
1. LEARNED
It's interesting how the smallest clusters were related by pathway and function while the next largest clusters were based on spatial location of the species in the cells. I knew that pleiotropic genes are genes that have many different effects on phenotypic variation, but I learned that you can attempt to measure how pleiotropic a gene is by looking at the extent of its negative genetic interactions with other genes. These genes were often on the edges of the network map because they don't fit well into a single cluster. It was cool to see how helpful the annotated clusters in the overall network map were able to show so many different hierarchies of location, function, pathway, etc. It seemed logical that most of the essential genes had more negative and positive genetic interactions than the nonessential genes and that the essential genes were more intertwined in the overall network.
2. MOST PRESSING QUESTIONS
Why would a lot of the positive genetic interactions be between genes in very different locations or different cellular functions? What in-depth studies would be needed to analyze these positive interactions?
How can we apply this data to the network mapping of the human genome and proteome?
How to the DamP and TS alleles work to knock-out genes?
3. PRESENTATION TOPICS
Explanation of the supplementary figures.
Explanation of mathematics used to test for interactions and significance.
4. THOUGHTS
I'm super impressed by the extent of this study. Creating that many single and double mutants and then studying them all and then analyzing the results is an extensive and time-consuming process. I love this paper in terms of its ability to look at the overall network maps and then to narrow down to figure out what pathways orphan genes belong to. I'm curious what technologies we could think of in order to speed up the experimental methods in this study so we can move on to studying things like triple mutants and the mutants' responses to chemical and environmental perturbations.
0 Knew
ReplyDeleteFamiliar with various methods of network reconstruction discussed in class (nodes=genes; edges=gene pairs that share similar profiles). That our ability to interpret inherited genetics is limited by the robustness (overlapping processes and functions which render most genes dispensible) of genomes. Familiar with use of TS mutants to assess cell cycle arrest in yeast/use of epistasis analysis to determine order of gene function from cancer biology course. Why/how you'd use Pearson correlation to assess these networks.
1 Learned
.Definitions of negative and positive genetic interactions (feel like I probably learned that before in cell bio, since it seems pretty fundamental-but forgot since then). Also learned that (-) interxns occur more frequently.
.As would be expected, essential genes have more interxns than non-essential genes (feel like this is obvious/was probably stated in a previous network paper, but this paper solidified it to me)
.Definition of phenotypic capacitors (=class of genes whose inactivation may incr. phenotypic variation among genetically diverse individuals in a pop.); highly pleioptropic genes (i.e. phenotypic capacitors) exist in the sparse areas of global networks, whereas functionally specific genes exist in the densely connected regions (which is how one may predict function based on location for such genes).
2 Pressing
What types of genes are inclined to have low degrees interaction (as this "suggests that these genes are under reduced evolutionary constraints and subject to condition-specific regulation")? [pg aaf1420-7, column 3, paragraph 1]
Why are there more (-) interxns than (+)? What does this have to do with LOF mutations? What is the evolutionary significance?
3 Presentation
Explanation of figures
Genetic networks involving gain-of-function alleles
4 Thoughts
It was nice to see the contrast in my understanding of network analysis, reading this paper in comparison to those earlier in the semester!
Stephen Lee
ReplyDeleteClass 26, Assignment 25
A global genetic interaction network maps a wiring diagram of cellular function
(0) Knew:
I’m familiar with studying genetic interactions using S. cerevisiae as a model (and how these interactions can be used to flesh out genotype-to-phenotype relationships) and some of the genetic phenomena the authors describe, like pleiotropy. I am not familiar with the methods the authors used to conduct their pairwise interactions.
(1) Learned:
How is it that the global similarity network so well isolated functionally-specific genes from pleiotropic ones (Fig 4 onward)? What, if any, is the relationship between the class of gene (TF, kinase, etc) and its likelihood to have a pleiotropic quality? How is a positive interaction qualified in this paper?
(3) Presentation Topic:
Exploring the methodology in greater detail (assembling interaction networks using double mutants) -- review supplementary material
(4) Thoughts:
This paper was generally confusing to me (certainly not accessible to an unacquainted audience), hoping reviewing it in class will shed some insight.
A global genetic interaction network maps a wiring diagram of cellular function
ReplyDelete0. KNEW:
I knew about basic transcriptional biology including genes and mRNA. I had seen genetic mappings of functional interactions before. I knew what Pearson correlation is. I was aware of clustering methods.
1. LEARNED:
I learned that Synthetic genetic array (SGA) analysis is a method of testing lethal genetic interactions. I learned that this is done with S. cerevisiae and can be used to gain functional information about gene interactions. I learned about the query genes are. I learned that positioning within the diagrams was importantly related to function because different but related processes were close together. I learned how the networks created could be used to predict novel gene functions.
2. PRESSING ?:
Can synthetic genetic array work with human cells?
3. PRESENTATION:
I would try to break down some of the tricks used to access the networks.
4. THOUGHTS:
This paper was hard to read and had a lot of jargon related to methods used.
Yes it can be - yeast has far fewer genes than humans do, so it was easier to implement. It should be relatively easy to extend it to human cells.
Delete0. Knew:
ReplyDeleteUsing correlation coefficients to define whether genes are "similar" or not and mapping into clusters to define a network map.
1. Learned:
The spatial analysis maps were interesting and seem to provide a lot more visual information than the global network profiles. Positive and negative genetic interactions... still a little bit unclear how these are defined though. The paper also provided insight and a good example of how the network mapping we've been learning all semester is applied which is something I've been struggling with.
2. Pressing ?:
"Spring-embedded layout algorithm"? Type of SOM?
3. Presentation:
NxN vs ExN vs ExE
4. Thoughts:
The number of authors... wow!
Agree with other posts that it was a very dense article; however, from what we've discussed in class I was able to follow the overall goals of the paper.
Class 26. Assignment 25. M Costanzo et al., A global genetic interaction network maps a wiring diagram of cellular function.
ReplyDelete0. KNEW:
In the other bio course I'm taking, we had a few weeks-long module on genetics which discussed the typical genetics approach of knocking out a gene to see what it does, and how yeast is typically used as a model organism (because it's easy to obtain and grow, and you can make it behave in a haploid fashion).
In class, we've talked a lot about finding ways to get an informative birds-eye view of how the different micro pieces of biology are working together to make a typically complicated macro structure. For example, in this paper, we are going from the micro view of an organism consisting of a bunch of genes to the macro view that clumps of genes represent interacting celluar components or processes.
Exactly how to go from the micro picture to the macro picture is the million dollar question. One way to conceptualize this transition is by connecting all related genes to make a gigantic network, after which you can do some further analysis to assign some meaning to clumps of related genes. This is the so-called 'wiring diagram' referred to in the paper---something that gives us an idea of how all the moving parts are interacting with each other.
To me, the paper represents a concrete realization of that dogma---going from micro to macro and representing what you learn, all through the construction of a network. This concrete realization is done, of course, in everyone's favorite model organism (Saccharomyces cerevisiae), because it is the organism whose genetics we know the most about and can most easily control.
1. LEARNED:
To be honest the main thing I learned is that people did this analysis. See the previous part (0. KNEW) to see how I think this relates to everything we've learned in class.
To repeat myself a little, I think it's mostly a group of people putting to practice our plan of making a network of genes to learn something about how they work.
2. QUESTIONS:
Given the advent of things like CRISPR, how easy/fast will we be able to do an analogous wiring diagram construction for some more complicated organism like humans? How informative would such an analysis be?
3. PRESENTATION:
I like to say that broad ideas are more important than technical details, and Dr. Wikswo's comment above seems to indicate that he agrees with me as far as this paper goes. I think a good presentation of this paper would briefly mention what they did here, and then indicate all the ways in which we can frame the paper in terms of what we have discussed in class.
4. THOUGHTS:
This is the end...I am a little sad. I think we've done well, though, and learned a lot this semester. At least I know I have.
(For some reason, it never occurred to me to post a PCRC on a paper I presented on...)
ReplyDelete0. KNEW
I had a basic understanding on how machine learning methods such as SOMs were used to rearrange data into useful formats for determining orphan gene function. I was aware of the redundancy of genes, also.
1. LEARNED
I never really thought about how multiple mutations would lead to greater numbers of lethal interactions. To do this, the researchers had to use thousands of mutated knockout yeast strains and crossed them with thousands more heat-sensitive strains to obtain the positive/neutral/negative interaction result. Because their outcome was dependent upon yeast fitness, it served as a good screening tool for potential drug targets. The manner in which genes interacted with other genes gave insight into how each gene was related; after making the cluster analyses, GO analysis was conducted to label them.
2. QUESTIONS
It would be significantly more difficult to conduct this experiment on human cells, but I wonder by how much? Or if there is reason to conduct this on different cell types?
3. PRESENTATION
Explaining the figures was the best part, I believe, followed by explaining the methods. Knowing how the figures were made helped a lot in tying it back to concepts covered in class.
4. THOUGHTS
A fun paper! I'm glad the class liked it; I never would have guessed it worked so well with what we've discussed since I mentioned it in Slack. I've learned a lot and am glad I was able to participate in this class.