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.
Kelly McGee
ReplyDeleteClass 06, Assignment 6a
Cisplatin in cancer therapy: molecular mechanisms of action
0: What I Knew: I was aware from our previous discussions of cisplatin of its chemical structure, its carboplatinum analog, its general use as a solid tumor chemotherapeutic agent, its method of creating DNA adducts, certain elements of modes of resistance, etc.
1: What I Learned: I learned that the drug paclitaxel targets cell microtubules, and I learned about the existence of photodynamic therapy. I learned that cisplatin is inducted into the cell by a copper transporter which is then degraded by cisplatin, contributing to cisplatin resistance. I learned that cisplatin activation of the ERK MAPK module may or may not encourage cisplatin resistance but activation of the p38 module is correlated to cisplatin effectiveness.
2: Questions: What are some experimental protocols for discovering and locating these correlations? As we discussed in class, individual gene knockout studies are often if not always insufficient.
What's the state of photodynamic therapy in clincal practice?
3: Presentation Topic: Photodynamic therapy, copper transport proteins and methods of degradation related to cisplatin.
4: Comments: Easy to read article for its size. Maybe because we already discussed the previous cisplatin article.
0-1) Good
Delete2) We will have to discuss the generic mass-spec protocols, but I'd prefer that we simply accept what they measure for now.
3) ;
4) Good - it's a big article.
0. Knew:
ReplyDeleteI was unfamiliar with this project before the beginning of the semester. I knew that there were problems with chemotherapy resistance (and other drugs) and that discovering these mechanisms was critical. I had not heard of such an approach to investigating the cause of these problems.
1. Learned:
I have a MUCH greater understanding of RTA after reading this paper. I learned about the processes in much greater detail than I understood before and understand how this approach can be expanded to other applications as well. I had a pretty good understanding of the 1st phase from class because I am more familiar with these techniques. The 2nd and 3rd phases make a lot more sense to me now after reading. I was very interested to learn of the calcium and estrogen dependence on Cisplatin resistance. I know from previous bio and physiology classes the importance of Ca in cellular mechanisms and was not surprised to learn that it plays a role here as well.
2. Pressing ?s:
Most of my questions were about why specific things were chosen to add to the study (see 3-2). I think just knowledge of what these are would help me to answer those. One thing that I could use an explanation for is what exactly is meant by a “seed” for the networks. Is this just a piece of known knowledge that can be input and then the database is used to find paths that lead to and go from this seed?
3. Presentation:
There were a couple of common themes that I think could make for good presentations topics:
1) Caspase cascade – what is it and what are its downstream effects?
2) HUWE1 – it was slightly unclear to me why this was chosen. I think a better understanding of what it is would help me to understand why it was discussed in the paper.
4. Thoughts:
I’m really glad this paper was shared. The in-class discussion was somewhat rushed and I didn’t have the necessary background at the time to fully understand and appreciate this project. As I said before, I have a much better understanding after reading this paper and was able to follow along better than some of the previous papers. The paper is very well written- I like how it is formatted and that a lot of information is in the supplement.. I think this made it a lot easier to grasp and then go and get the details later.
0-1) Excellent
Delete2) Ask James to explan.
3) <>??>>
Ask Nicole about HUWE1.
4) Glad you like the paper. We will try to get James to return to discuss inverse network analysis in a more formal manner. For now, the trick is to see how a giant MOA experiment works under RTA. When I get back, I want to hear what Zach has to say against RTA!
0. Knew:
ReplyDeleteI knew the general overview of this project from Nicole’s presentation done previously in class. I knew of the goal of the project to identify the MOA of a specific, possibly even unknown, compound. I knew that there were three main phases for the project, and I understood that an expanded canonical network was produced for cisplatin from this project
1. Learned:
I learned more about the HUWE1 pathway and why it was selected for further analysis. This was either not fully talked about in the previous talk, or I did not notice/remember it (my apologies if this is the case). This pathway was chosen in an attempt to compliment the data driven network by identifying events that occur outside of the MOA of cisplatin. The HUWE1 pathway was beneficial for further analysis because it is one of the top five raked hits and had many fold changes over time.
2. Pressing Questions:
It was mentioned that this 30 day process should be able to work for unknown compounds. The conclusion mentions this would be beneficial for threat assessment unknown chemical/biological weapons. This led me to this question: if a MOA or response/resistance network is determined from an unknown compound, would it be possible to determine what this compound is from just this information alone?
Another question I have is this: are there any types of compounds that this project would not successful with?
3. Presentation Topic:
I would like to have a more in depth presentation on the network construction, such as the tools used and the procedures performed.
4. Thoughts:
I think this is a very cool project, considering that analysis of this kind before the project took way longer than 30 days (I think it 18 months?). I am looking forward to hearing more about this project and its outcomes in the future.
0) Good
Delete1) I would like Nicole to discuss the method by which HUWE1 was added.
2) This is an interesting question. My view is that you don't give a white powder to an analytical chemist without expecting an immediate analysis. However, there are good examples of unknown designer drugs that caused illness by unknown mechanisms - check out
https://www.cdc.gov/mmwr/preview/mmwrhtml/00000360.htm
As to <>, we need to have that discussion later in the semester, with Melissa!
3) Forthcoming, but not tomorrow.
4) We will discuss bendamustine once the analysis is done.
0. KNEW:
ReplyDeleteFrom class discussion I knew about the growing analysis of cellular mechanisms and how this knowledge is being used to build an understanding of biological networks. I had also previously learned that the goal after 30 days was to determine a mechanism of action, and that this 30 day process was divided into three distinct phases.
1. LEARNED:
I did not know what a multi-omics platform was, but quickly learned the importance of what a standardized database could accomplish through the comparison of data. I was aware of the concept of such a platform, but not the specific terminology. I was unaware that this procedure could be applied to uncharacterized compounds, and that is why previous experimental data is not used.
2. PRESSING QUESTIONS:
It is stated that future work will harness technological advances. Will updated technology help enhance the performance of this technology platform or decrease the 30 day scale over which it collects data and analyzes the mechanism of action? Also, the mechanism of action of cisplatin is described as being well understood, so would any extra work be required to determine a previously unknown mechanism?
3. PRESENTATION:
The "diverse applications" of this platform. This topic may be difficult for someone with little experience on the subject, but the examples in the paper are interesting (pharmaceuticals, chemical or biological weapons, pollutants, etc.).
4. THOUGHTS:
The first time we heard a presentation about this topic I was overwhelmed. I am glad this paper was shared so I could have a better understanding and appreciation of the work.
0) OK
Delete1) Multiomics is a real big deal.
2) Yes for unknown. As to new platforms, we will hopefully get to mummichog... Not sure that there will be time.
3) We need to think about this a bit more before trying to design a presentation.
4) I'm glad that the second pass is not as overwhelming. Look at what you've learned!
Chinowsky_TheorSysBio_PCRC_09122016
ReplyDelete0. Knew
From Nicole’s presentation earlier in class, I understood that RTA consisted of three phases, and used a combined –omics approach (proteomics, metabolomics, and transcriptomics) to rapidly assess the many changes that occur at a cellular level as a response to a drug. It is the thorough analysis of the cellular responses that allow for rapid development of a MOA. For cisplatin in particular, several expansions to the canonical model were made, including an estrogen dependent response.
1. Learned
This paper further broke down the three phases used for RTA, and I felt like I now have a much better understanding of what constitutes each phase, and how the multi-omics approach allows for rapid assessment and data-gathering. I see now that phase one is integral, particularly for new compounds that have not been used in a clinical setting, something that was not apparent to me from the presentation. I also feel like I now have a greater understanding of how the generated data sets are used and transformed into mechanisms. This is made particular clear by how the HUWE1 pathway was selected for further analysis in exploring novel secondary mechanisms.
2. Pressing
Are there any situations in which this model of RTA would not be sufficient/useful? The end of this paper mentions that this platform might be useful for rapid assessment of unknown chemical or biological weapons. However, if it is unknown what organs/cell types these weapons affect, where is the starting point for RTA?
3. Presentation
How one might be able to integrate imaging approaches to detect mechanistically important events, i.e pg 24.
4. Thoughts
This paper made some of the more convoluted/rushed parts of the in-class presentation more clear. RTA is a very interesting idea, and I’d like to know how it might be applied in the future.
0) Good foundation from previous class
Delete1) Excellent!
2) See my responses above. MPTP is a perfect example - it hurt a goodly number of people. That type of problem continues, and may be getting worse. Imagine if done by a hostile group!
3) <>
4) Good, and yes.
Ben Terrones
ReplyDeleteSysBio16 Asgn_6A_Class_06_Article_06_2016_09_13
0. KNEW:
I only knew what was presented in the lecture on RTA from a few weeks ago. This includes the basic process involved in RTA, which is divided into three phases. In phase one the dose and exposure time is decided; phase two utilized different methods to measure significant changes in species, and in phase three this information was used to construct the mechanism of action of the drug. They were also able to compare their empirical network to the canonical mechanism of action and then develop a data driven network from the empirical data.
1. LEARNED:
This paper really helped me understand the specifics of the RTA process that were not as clearly explained in the presentation because of time constraints. This included the process of how the optimal dose was selected. I didn’t realize they found the optimal dose in such a way that they could use the same procedure even if they did not know the identity of the compound. It was also interesting to see how they used the data to find and construct resistance mechanisms, specifically estrogen-induced resistance.
2. PRESSING ?:
For other drugs/compounds where there is not as much information in the literature, how would they validate their results?
3. PRESENTATION:
How this technology could eventually be used to provide personalized medicine for patients and other potential uses of this technology.
4. THOUGHTS:
This paper really made me realize the promise that RTA has and really the promise that all combined approaches have. The most interesting application that was mentioned in the paper, to me, is the potential to assess infectious diseases and possibly find treatment more rapidly.
0) OK. We had to start somewhere...
Delete1) Optimal dose bit is neat, but there is a problem if a drug has different high- and low-dose MOAs (which does happen).
2) Let's watch bendamustine. In actual practice, validation would be even more important.
3) Not yet.
4) Good!
4) potential to assess effects of cellular intoxication is an interesting application, but could figuring out why drugs in pharma are failing be more important? If you find the why's, the treatments will follow.
DeleteSylvia Morrow
ReplyDeleteAsgn6A_Norris, et al.,Global cell response captured in 30 days yields comprehensive drug mechanism
0. KNEW: I recognized the majority of the big picture from Nicole's presentation in class.
1. LEARNED: The section titled 'Capturing the dynamic processes that govern cell fate' (starting on pg 14) was very informative. If I understand correctly the initial analysis just matched database entries with species that had statistically significant changes with time. This section talked about using the order of the time point data to look at causality relations which seems like an important addition, although I have the general sense that complex feedback mechanisms make causality a little less important than I'm used to considering it.
2. PRESSING ?:
--Is there a normalization process that accounts for the fact that "significant change" will mean something different for each species? Some sort of comparison to a control?
--What is preventing this type of analysis from becoming a common practice? This leads to secondary questions about what it means to have done this in 30 days: how many people were working on the project 24/7? who was supporting this research? what would change about the process if you, for example, doubled the time period? If this type of analysis becomes common practice will might that alter the way people decide what type of data collection they're going to do?
3. PRESENTATION:
--Overview of KEGG database (or bio databases in general) to understand things like what type of information is being gathered and what the level of reliability is in order to better understand the potential queries.
4. THOUGHTS: There were a lot of very specific information in this paper that made it a bit tough for me to wade through, but I found the big picture pretty accessible.
0) OK
Delete1) Causality may be hard to infer with feedback - an important point. There were not enough time points for this experiment, although bendamustine is a bit better.
2) Typicaly, the MS data does not provide absolute numbers without calibration.
Nicole knows the people count. Were the time and budget doubled, one could look at different cell types, more time points, or long-term responses after short doses, instead of the continuous doses used. This leads to the important problem of optimal design of experiments - what is the best experiment to test two conflicting models or test a specific hypothesis. This would drive data collection strategies - we should look at it (Ross & the Adam robot). GOOD POINTS.
3) <> Ask James Pino for the list!
4) Yes, good.
2) our significant change criteria stays the same for all cell types. HL60s gave significantly less significantly changing species, but we expected that. When running a long-term experiment, you want to change as few variables as possible in order to try to gain comparable correlative data
Delete- 35 people, 30 days, a lot of man hours.
- Research supported by DARPA
- Honestly, the protocol wouldn't need to change to 60 days. We get ample time points and sufficient data collected in 30. The only thing another 30 days would allow for is more cell lines. For bendamustine, we ran 12 time points across 6 modalities (2 metabolic, 3 proteomic, and rna seq).. Due to budgetary constraints, RNA seq was only ran for 4 time points. Due to lack of instruments and time, proteomics was only able to run SILAC on half the time points.
If we wanted to become more efficient in a 60 day period, our best bet would be to train more people to analyze the data. This is our weakest link. It's very tedious, manual analysis traipsing through 750,000 rows of data.
Natalie Hawken
ReplyDeleteReading Assn 6A, Article 06: Global cell response captured in 30 days yields comprehensive drug mechanism
0. KNEW
From the first class article (Siddick), I knew the basics of the cisplatin MOA and the resultant methods of resistance in the cancer cells. This knowledge matched with the article's identification of the canonical model. From Nicole's initial RTA presentation, I knew the three phases of the process, the basics of the expanded canonical and data driven networks. Nicole had also mentioned briefly the role of estrogen in cisplatin resistance.
1. LEARNED
I was really interested to find out how the target dose and exposure was selected. Figure 2 helped to clear up just how the specific dosages were scientifically picked for maximum, legitimate results. The supplemental information on experimental procedures helped to clarify why each proteomic/metabolomic/transcriptomic platform was used. I am still a bit confused on how they all work. I learned how they used multiple sources of species' change in activity to determine which were actually significant to the MOA. I hadn't heard of the HUWE1 pathway to apoptosis, and it was interesting to see how this pathway normally leads to apoptosis, but can be switched over to an anti-apoptotic path. I also learned about ATP1A1 pathway, STIP1 cascade, and other mechanisms of resistance that hadn't been presented in the Siddick paper.
2. MOST PRESSING QUESTIONS
On pg. 13, the author notes that a full MOA is limited by the amount of data in the KEGG database. As this database has hopefully grown or put in combination with other databases, has the DDN for cisplatin expanded? Are any of the previous disregarded species changes now implicated in the MOA?
In the methods, its stated that the cell line used (A549 cells) consists of human lung carcinoma cells. Given that cisplatin works very well for certain cancers (like testicular cancer) but can be very ineffective for other cancers, how would the multiomics data vary for different types of cancerous cell lines tested? Could the results from different cell lines even be accurately compared to each other?
3. PRESENTATION TOPIC
Guide to multiomic platforms: The experimental procedure (pg S-15) discussed the general overview of platforms like label-free, SILAC, phSILAC, RPLC, etc. But I would like a presentation on how each of these tests are run and what sort of results each specific platform produces.
Explanation of novel pathways (see figure 5): I would like a short overview of the additional pathways discovered through RTA including the CUL4B/HUWE1 pathway, the ATP1A1/Ncx1 regulation, the estrogen-induced activation of mTOR, and the STIP1 cascade. I would like to be able to fit these pathways into the Siddick model to determine how all of the apoptosis/anti-apoptosis paths work in tandem.
4. THOUGHTS
I found it much easier to digest the RTA information in this paper as compared to the talk given in class. Topics that were glossed over in lecture (like ECN v. DDN, multiomic protocols, additionally found pathways) were much more in depth in the paper. I'm really interested to see the changes made in the process for the bendamustine 30-day challenge, and I'm curious to look into the bendamustine results for novel pathways.
0) Good.
Delete1) Well stated.
2) Ask James Pino re extent and extension of KEGG annotations. In effect, our paper should lead to the expansion of KEGG, but I'm not sure how that will occur. Possibly not until validation experiments are conducted.
As to other cell lines, right on.
3) <> Other pathways - ask Nicole in class.
4) We are all interested in seeing what is learned from Bendamustine
2) To clarify, a full MOA is limited to the interactions annotated in KEGG, not by the annotated species themselves. The species can be in the database but not well known enough to have its interactions clearly annotated, specifically protein-DNA interactions.
Delete3) specific multiomic platform procedures are outlined in the very large supplemental material packet coupled to this paper. Scientific journals are like twitter.. character count is very limited
0. KNEW
ReplyDeleteFrom previous discussions and readings, including a helpful visual of the Pino model, I had an understanding the scope of complexity regarding intercellular mechanisms of action. I recalled some of the functions and behavior of cisplatin, including specifically the interaction of its platinum group on DNA adducts, and how releasing cisplatin in a cell initiates a variety of pathways downstream.
1. LEARNED
An interesting phenomenon that was presented early in the article is that of pleiotropy, in which a primary MOA introduced exogenously to the cell initiates pathways that have seemingly totally unrelated results. Upon a little further research, I found an analogy of one gene of DNA influences several seemingly unrelated phenotypes that are expressed.
I also learned about approaches increasing efficacy in the analysis of multifaceted cellular response mechanisms. A multi-omic approach is found to be most useful in molecular profiling when considering specific cellular reaction to an introduced compound. I have a much greater understanding of the methods discussed in this platform for capturing MOA beyond gene regulation (I worked on a project involving somatic recombination this summer).
2. PRESSING?
Perhaps an obvious question, but what can be done to decrease incubation analysis times in order to more rapidly assess the pathways affected by the test drug? Can anything be done to bring this idea to low-resource areas, as a sort of analog to an RDT?
3. PRESENTATION
An overview of KEGG (I find the predictive modeling very interesting), specific downstream effect of the p53 cascade that effects apoptosis
4. THOUGHTS
This paper conveyed its information concisely and descriptively, and my limited background knowledge in cisplatin was sufficient in getting me through the bulk of the paper unimpeded.
0) OK.
Delete1) Good to read up on pleiotropy.
2) Ask Nicole. We will reduce it to 100 ms.
3) <>, <>
4) Good!
Grrr - I just figured out that the conversion to HTML is messing up whatever I put between double < and double >.
ReplyDeleteThis comment has been removed by the author.
ReplyDelete0.Knew: I knew about cisplatin’s usage as a therapeutic treatment for cancer from a previous article we read on cisplatin. I was also familiar with the tendency of some cells to be resistant to the DNA-damaging effects of cisplatin.
ReplyDelete1.Learned: I learned about the importance of this project with respect to illuminating secondary mechanisms of action following cellular exposure to exogenous compounds. I learned about the role of each phase in the procedure mentioned and what can be observed at different time intervals in the course of the allotted 30 days. I learned that screening the cells is important to determine the optimal dose of the exogenous compound to be administered, and I appreciate that this paper gave the amount of cisplatin that, when applied, produces a maximum response and maintains cell viability (about 50 μM). At this amount, cisplatin’s toxicity can be observed as early as an hour after application. I learned about the relationship between calcium and cisplatin resistance; disruption of calcium signaling and the reduced expression of calcium-related potassium channels, according to the article, are implicated in resistance to cisplatin. I also learned that estrogen and ATP1A1 are related to cisplatin resistance.
2.Pressing Questions: The article mentioned the revelation of a cisplatin-resistant population of cells in the study, but it does not really explain how those cells are different from the ones in the population that have succumbed to the effects of cisplatin (p.16). The article mentions that developments in network analysis “provide a foundation for identifying and mapping non-linear interactions and associations not detected by reductionist approaches.” What are these reductionist approaches? (p.6). Why were A549 cells specifically used for this study? How are they more advantages to use than another cell type? (p.7). What is the Qiagen IPA causal network analysis tool, and what are some more of its applications? (p.14).
3.Presentation: more about IPA, and maybe a simple explanation of how HUWE1 modulates DNA damage response and apoptosis pathways.
4.Thoughts: One of the things I did not like about this article was that the figures were not shown where they were referenced in the paper; seeing the visual representation and the written content together would probably have helped enhance understanding. Additionally, it seemed odd to me that the “Results and Discussion” section was in the beginning of the paper, but other than that, the organization was easy to follow.
1) Good!
Delete2) Think about the resistance like bacteria. What makes antibiotic resistant E.coli different from regular E. coli?
-Will get to more network analysis later
-Ask Nicole or James about A549 and analysis tools
3) Ask Nicole about HUWE1
4) Manuscripts are often formatted different from a published paper!
2) A549 cells were used because that's what the biologists on the project specialize in. We've now transitioned to HL60 suspended, immune cells because we wanted to use a suspended cell line... and the biologists are specialized in using them.
Delete3) IPA is a great tool. Pharma uses it frequently. Not so great for scientists publishing work as it is not open-source. very expensive.
4) The layout of the paper was necessary for cutting ~50,000 characters from the original rough draft.
4) But I appreciate your critique. These are necessary from readers in order to improve writing.
Delete0 Knew
ReplyDeleteI was somewhat familiar with the RTA approach from our class discussion.
1 Learned
I thought the dose optimization approach in the first phase was interesting; I hadn't fully appreciated the consideration that went into assay optimization before the data acquisition phase. The resistance work was also very neat. At the onset I wondered how a population based assay would be able to identify mechanisms of resistance which are specific to small subpopulations, but the ATP1A1 and NCx1 work was interesting.
2 Pressing Questions
What does it mean to "seed" the empirical data set with the canonical mechanism? How does the Qiagen IPA causal network analysis tool work?
3
What does it mean to seed a network?
4 Thoughts
How can we move from identifying hits as "important" to creating causal mechanisms of action.
1) I agree! I can only fathom how much work it is to optimize in such a short period of time.
Delete2) Ask James today.
3) We will get to more network analysis later in the class
4) Ask Nicole and James about this process.
Seeding a network:
Delete- Identify species of interest
- Drop species as .csv into James' Python script
- script queries Kegg, Reactome, and HMDB to identify pathways that the species of interest are involved in
- This leads to a built network
"Expanded Canonical" was seeded with all canonical species.
"Data driven" was seeded with all significantly changing species from our dataset.
Qiagen IPA allows you to pick a pathway of interest, i.e. apoptosis. It then uses your temporal data to show you what species were detected at the individual time points in the data. I can show some screenshots in class to give a better explanation.
In regards to your questions, I thinking seeding a network is establishing base nodes and I think the Qiagen IPA causal network analysis tool works via data sources that are probably public but not easily identified via the proprietariness of the software. I think these are both things we should find out for the construction of our presentation (not the presentation itself).
Delete0 Knew: Cisplatin is a cancer drug that causes damage to DNA as well as a basic understanding of RTA from the class discussion
ReplyDelete1 Learned: Gained a better understanding of the method for optimization of dose and exposure.
2 Pressing Questions: What if any plans exist for the conversion from a fuzzball 3D depiction of interactions to a list of known effects the drug has on the cell?
3 Presentation: A quick look at KEGG and how it classifies its information
4 Thoughts: The paper helped reveal too me the incredible market for RTA.
From the cisplatin side of things I felt like it was easy to get through and understand especially with the cisplatin presentation in class as background information.
Stephen Lee
ReplyDeleteClass 06, Assignment 6a
Global cell response captured in 30 days yields comprehensive drug mechanism, submitted
(0) Knew:
Based on Nicole’s presentation on the topic, I knew that cisplatin was an appropriate test drug for RTA due to the body of literature describing it (canonical mechanism), and that high-resolution mass spectrometry was used in the project to develop a “rapid proteome screen” via scoring over given timescales to evaluate molecular response.
(1) Learned:
I learned about the logic and methods behind investigating the HUWE1 regulatory circuit as part of the data-driven network, and the effectiveness of RTA at identifying this and other species that were not represented in the canon, but were significantly perturbed during dosage analysis. I also learned about RTA-mediated elucidation of estrogen-induced cisplatin resistance via ERK.
(2) Pressing Questions:
Almost all of my questions regarding RTA concern discovery phase methodology, especially with regard to the molecular screen. In reading and thinking about this paper, I’ve essentially ignored my lack of knowledge concerning the methods presented, or tried and failed to learn about these techniques in any fulfilling detail. One aspect of Nicole’s presentation and this paper alike that both impressed and confused me was the number of data points collected. What do each of the ~250k data points and ~54k unique “species” represent? Can ~13k different statistically significant molecular changes really be identified in one assay? How are there even that many unique species? What instrument is gathering these data and how? Hopefully I can put some questions forward in class that clarify my concerns further.
(3) Presentation Topic:
A presentation succinctly and specifically describing high-resolution techniques that contribute to the immensity of data gathered during the discovery phase (like MALDI-MS) to evaluate molecular perturbations in the samples.
(4) Thoughts:
Appreciated further clarification of RTA and am looking forward to the in-class discussion.
I too am perplexed by the amount of data generated. What I got was that transcriptional data through RNA-seq and protein data via some kind of MS method generated all the entities. I assume each molecule has a 'signature' by these and that this is matched in respective databases? What I got is that this was done with KEGG. May be wrong. I imagine if it is different, it is located somewhere in the methods towards the bottom. What gets me, is there not noise when measuring these? How can this be mapped accurately to an identification dataset? The significance thing as well is mind-boggling. Those are all found from comparisons with controls, correct?
Delete0. KNEW:
ReplyDeleteI knew basic cell biology and pathways that seemed required to understand what was being mapped. I understood what mass spectrometry is and how this is related to the dose determining steps. I am familiar with what transcriptome, proteome, phosphoproteome, and metabolome data would theoretically include and what their respective fields of study are. I knew about databases, although not exactly what they contain. I knew about graphs and networks.
1. LEARNED:
I learned what caspases are and their role in programmed cell death. I learned what RNA-Seq is. I learned what matrix-assisted ionization and why it is useful (necessary?) for this application. II learned the importance of having similar significance percentages between the ECN and the canonical MOA dataset. I learned about the Qiagen IPA causal network analysis tool. I learned the primary pathway of data manipulation: empirical MS and RNA-seq data -> identified partially by KEGG -> assembled into two networks (DDN on unique significant compounds, ECN with canonical data). I learned what the DDN, ECN, and ERN are. I learned an overview of the technical side of the network analysis.
2. PRESSING ?:
Supplemental Figure 2: How was the cell viability and thus the dose conceptualized chosen?
Line 272-275: How are these values related? Where does the DDN size come from? More not understanding the values than anything. (Update: makes sense with figures).
What could be done with the data that isn’t in Kegg? How ‘complete’ is Kegg in regards to what is known? How is KEGG maintained?
3. PRESENTATION:
I would split the presentation on the three main sections of the process. I would emphasize the network analysis, using examples from the generated testable hypotheses.
I would absolutely do a short presentation on KEGG.
4. THOUGHTS:
This was exactly what I wanted to read in regards to the RTA. Definitely feel a connection for all what we have learned.