Complex Systems Discussion Group

 Forgot Password
 Register now

QQ登录

只需一步,快速开始

Search
12Next
Return to List Add thread
View: 20541|Reply: 32

[Challenge Topic] Evolving out of cancer (Matt Thomson, Simone Bianco, Zhiyuan Li)

    [Copy Link]

43

Threads

186

Posts

524

Credits

Moderator


IP  162.105.160.10

Posted at 2012-4-30 11:59:34 | All floors |Read mode
Edited by zhiyuanli at 2012-5-7 23:07

Original Link: http://www.ucsf-pku.org/index.php/Evolving_out_of_cancer

Challenge Statement
The rapid emergence of drug resistance in response to therapy is a major current challenge in cancer treatment. We are developing increasingly powerful therapies targeted to the genetic lesions that drive particular subtypes of cancer. However, in almost every case, tumors quickly develop resistance to drug treatments, and this currently vastly limits the scope and power of molecular therapies. Drug resistance is a challenging problem because it is a complex evolutionary process operating at many levels of organization including signaling pathways (signaling pathway dynamics), evolutionary dynamics of cells, organ environment, so that tumors can evade a drug using mechanism ranging from changes to signaling pathway dynamics to changes in microenvironment. The current challenge is to develop computational approaches for integrating the large amounts of data from sequencing of tumor samples, population level and clinical trial data on patients and lifespan following treatment, molecular data on signaling pathway and interactions, biochemical data on drug targets into new strategies for over-coming the emergence of drug resistance. The goal of this challenge is integrate concepts from systems biology and evolutionary biology with computational tools and high-throughput data to propose new approaches for combating or evading the onset of cancer drug resistance.

Approaches we will introduce Dynamical systems Modeling of pathways, bioinformatics with sequence and gene expression data, evolutionary dynamics, Genome wide association studies, epidemiology.

BackgroundHow specific to be? key question is whether to focus on a single specific type of cancer. Gleevec, Iressa background on known mechanisms of drug resistance.Give background on following topics: 1) Drug Resistance and known mechanisms. 2) Current progress in high-throughput data collection.

Outcomes:
Produce a methodology or novel computational approachPossible outcomes of this project can be summarized into two distinct research lines

What kind of data is needed to address the problem:  The data available now falls into several categories. Sequencing of cancer data is more and more frequent to try and understand the genetic differences between resistant and non-resistant tumors. At the level of the individual, highlighting genetic differences becomes fundamental to evaluate possible therapeutic outcomes. Moreover, the increasing availability of the data may also increase the epidemiology of cancer at a population level. On the other hand, cancer development and spread is fundamentally a micro-evolutionary process. This means that, under environmental pressure, cancer cells tend to be selected because of their higher fitness. Chemotherapy and radiotherapy represent strong selection forces, and escape from these forces selects “strains” of resistant cells. Evolutionary trajectories of cancer cells can help pinpoint when and how resistant mutations occur during a therapeutic effort. Finally, cancer development involves modification of important molecular pathways. The increasing amount of data available on changes in cellular pathways is increasing our understanding of the modification that occur at a molecular level. Ideally, linking possible molecular changes to the resistant phenotypes would create the opportunity to address the issue. Also Epidemiological data

What kind of modeling is needed to address the problem: The approach we want to follow is integrative. Data mining and knowledge must be integrated with a modeling effort to explain the data and predict possible outcomes of targeted intervention. Driven by the data available, we identify three different modeling approaches. A bioinformatic approach (please elaborate.) . A systems/synthetic biology approach, which should allow us to study how and why is the molecular network of pathways allowing drug resistance. On a larger scale, this approach may help understand which form takes the relevant molecular network, if there are “missing links” and how to deal with them, and, in general, how to map a system of this complexity to a dynamical system. An evolutionary biology approach, which would allow modeling of drug resistance both at the genome level (mutation) and at the cellular level (competition), might help discovering the genetic space in which cancer cells move, the fitness landscape that characterizes the dynamics, and the link between the fitness of the cancer cell genome and the affected cellular pathways.

Workshop schedule
Day 1: Tutorial session; brainstorming and narrowing down topics; write an outcome presentation.
Day 2: Study the problem, refine ideas; write a document/protocol white paper; more tutorials (if needed)

Possible homework:Exercise: Simple evolutionary biology simulation algorithms (small genome-wide simulations, rationale for binary genetic sequences), competition models (predator-prey model), simple network motifs applications (what are  feedback/feedforward loops, how do you achieve adaptation, how to model escape, etc.), some notions about large dataset analysis. some basis on simple software like MATLAB for bio-background students

Reading:Interesting review:Hallmarks of Cancer: The Next GenerationCell, Weinberg.
Selected chapters from Nowak’s “Evolutionary Dynamics” book and from Weinberg’s “The Biology of Cancer” book.
Selected chapters from Uri Alon's "An Introduction to Systems Biology - Design Principles of Biological Circuits".
给我一块二向箔,清理用

2

Threads

21

Posts

74

Credits

Moderator


IP  222.29.46.51

Posted at 2012-5-9 16:21:06 | All floors
I'm very interested in this topic!! But as far as I know, the cancer network hasn't constructed completely yet, so that the drugs designed now cannot solve the most fundamental problems and drug resistance occurs. So I bear an idea that maybe we can first construct the whole cancer network as complete as possible and then the remaining problems may be a piece of cake~~ This is my personal opinion, welcome to Pai

43

Threads

186

Posts

524

Credits

Moderator


IP  115.27.30.233

 Author| Posted at 2012-5-9 16:40:16 | All floors
Reply YuanYuan Add Thread

I like your ambition... yet constructing a full network of cancer is not a team challenge topic, it is a Nobel Prize project....

In this topic, we plan to focus on one specific network that is responsible for the adaptive response of breast cancer cell to drug (adaptation is good for E.coli, but can be very bad for treating cancer-- it means whatever you do, the cell ultimately go back to the fast-dividing state). I plan to figure out the topology and "weak point" of this system.

Do you have other specific cancer-related network in mind?
   
给我一块二向箔,清理用

3

Threads

13

Posts

115

Credits

Registered member


IP  162.105.21.111

Posted at 2012-5-9 21:33:49 | All floors
I'm afraid it's more than one Nobel Prize project... And I don't really believe a full network can help us understand cancer...

2

Threads

21

Posts

74

Credits

Moderator


IP  222.29.46.51

Posted at 2012-5-10 14:54:53 | All floors
Reply zhiyuanli Add Thread


    Not yet! But I am searching for some references about it. I will be very glad if you can tell me some good references

43

Threads

186

Posts

524

Credits

Moderator


IP  162.105.23.162

 Author| Posted at 2012-5-10 15:48:24 | All floors
www.nature.com/nrc/journal/v5/n5/abs/nrc1609.html: ERBB receptors and cancer: the complexity of targeted inhibitors

that's one of the biological readings I suggested in the breast cancer related system
and for modeling part, for sure, Alon's introduction to systems biology is the best introduction
给我一块二向箔,清理用

8

Threads

54

Posts

210

Credits

Intermediate Member


IP  162.105.160.33

Posted at 2012-5-10 16:00:26 | All floors
I'm very interested in this topic!! But as far as I know, the cancer network hasn't constructed comp ...
YuanYuan replied at 2012-5-9 16:21

Once I believed a complete network could resolve everything. But since last rotation project, I really doubt this method, at least so far...First, how do we know the network is complete? Are the prediction results sensitive to the network size? Sometimes the network construction is a little bit subjective...The second problem is, huge network is really difficult to analyze... That means we may not get useful information from it...On the other hand, does small network can explain something? It's hard to say...As for me, I don't believe simplified network is able to explain a complex disease such as cancer...

Reviews

Indeed we can break down huge networks into small ones. But it can only explain a local behavior instead of a global one. In nonlinear world, simply additive of networks shouldn't get additive result   Posted at 2012-5-11 21:29
indeed, small networks are insufficient to explain huge ones in many ways. but to "break down" huge network into small ones might help, as discussed in class last thursday.  Posted at 2012-5-11 11:12

8

Threads

54

Posts

210

Credits

Intermediate Member


IP  162.105.160.33

Posted at 2012-5-10 16:14:46 | All floors
Reply zhiyuanli Add Thread

I once heard a opinion that as long as one lives long enough, he must get cancer... The reason that incidence of a cancer is low is that before we could get cancer, we die... Statistically cancer is a senile disease.  So does longevity and cancer reach a balance?Will cancer cell get ageing?

2

Threads

23

Posts

113

Credits

Registered member


IP  162.105.160.10

Posted at 2012-5-11 00:33:46 | All floors
There are some mechanisms proposed by scientists,such as ATP-binding cassette (ABC) transporter superfamily which translocate a wide variety of substrates across extra- and intracellular membranes (http://www.ncbi.nlm.nih.gov/books/NBK3/). What I am interested is, is there anything in common between the evolution of cancer and the evolution of high resistance to antibiotics of bacteria?If they have some similarities, maybe the high resistance network of persister cells will help us understand the evolution of cancer, otherwise, these two mechanisms maybe two distinct evolution path.  

Reviews

indeed the mechanisms of cancer drug resistance are quite alike to bacteria resistance to antibiotics.both of them include the decreased uptake of drugs, cell efflux of drugs and alternation of metabo   Posted at 2012-5-14 23:12

1

Threads

28

Posts

146

Credits

Registered member


IP  162.105.87.183

Posted at 2012-5-11 11:07:13 | All floors
Reply QX_Cai Add Thread


    u always raise someting im intrested in~
there are 2 kinds of ageing: chronoligical & replicative. then, will cancer cell get ageing?
for a single cell, yes, cancer cells also age and die. but when it comes to replication,
unfortunately they dont have Hayflick limit, that means, they are eternally young.
What do u mean by "balance"?
王紫薇

2

Threads

23

Posts

113

Credits

Registered member


IP  192.168.1.141

Posted at 2012-5-11 12:15:08 | All floors
Reply


     u always raise someting im intrested in~
there are 2 kinds of ageing: chrono ...
Amber.Zw.W replied at 2012-5-11 11:07



won't the whole number of the cancer cells stay stationary?

Reviews

yes,if death is equal to growth rate, there will be a balance and a patient can live happily together with tumor. But actually growth rate is always much higher than mortality.  Posted at 2012-5-15 15:13

2

Threads

21

Posts

74

Credits

Moderator


IP  162.105.160.21

Posted at 2012-5-11 20:27:45 | All floors
Reply zhiyuanli Add Thread


    Thanks a lot! It seems I have underestimated the difficulty in constructing a full network.

2

Threads

19

Posts

109

Credits

Registered member


IP  162.105.23.227

Posted at 2012-5-11 20:37:40 | All floors
The main post stated that the current challenge is to develop computational approaches for integrating the large amounts of data. Recently, I read a paper published in the journal Cell [1], which presented an integrative personal omics profiles for a single individual. In their article, multiple data for the single individual, like whole-genome sequencing, transcriptomic, proteomic, metabolomic, and autoantibody profiles, were integrated. I think their way of integration may help us in this topic.

[1] Chen R et al. Personal Omics Profiling Reveals Dynamic Molecular and Medical Phenotypes. Cell 148 (2012): 1293-1307. http://www.sciencedirect.com/sci ... i/S0092867412001663

Reviews

we are all considering integrating or analyzing data at hand, but far from enough data are own, what can we do to probe the unkown from known? maybe we can think about that.  Posted at 2012-5-31 20:47
nice paper! thank u!~  Posted at 2012-5-12 21:02

8

Threads

54

Posts

210

Credits

Intermediate Member


IP  162.105.160.10

Posted at 2012-5-11 21:16:40 | All floors
Reply Amber.Zw.W Add Thread
What's the essential difference between chronoligical & replicative ageing? When I say balance, I mean cancer may competes with longevity. I guess there  exists a upper longevity limit in a given environment, above which everyone will get cancer. For species with good environment in abundance of nutrition, evolution will 'help' them reach the age bound individually and meanwhile make the whole population long-lasting. So once reach over this line, our biologically systems may get problems, cancer for example. In this sense, cancer is a evolutionary product. So how to evolve out of cancer? Maybe it is also a problem of how to live longer. Really challenging...Just like a competition with the GOD and nature..Personally, I prefer to grow naturally... I believe we can't be forever young, otherwise extra resources and energy are needed to maintain our life, but where is the extra resources? But for scientific research and curiosity, I am in...





1

Threads

28

Posts

146

Credits

Registered member


IP  162.105.87.58

Posted at 2012-5-12 20:58:12 | All floors
Reply QX_Cai Add Thread

replicative: cells have their limit for devision(termed Hayflick limit), they stop repllication after certain times of deviding.
chronological: when a cell doesnt devide, it still gets old due to the accumulation of all kinds of damage as time goes by, because no deviding doesnt mean no metabolism, just like a machine, it works, so it wears.
after reading some papers i gradually get the point that ageing is not a matter of TIME, it seems to be a matter of structual & functional degeneration. i agree with the "line", but it shouldnt be a certain age, it is just the extent of damage, because studies indicated that when the organism stays young and healthy at an old age, the susceptibility to cancer is also lowered ( colmam, 2009, caloric restriction delays disease onset & mortality in Rhesus monkey).

Is cancer an evolutionary product? it is a question the same interesting as whether ageing is an evolutionary product. perhaps both the answers areno. the article"p53 ancestry: gazing through an evolutionary lens" reminds me of this.
and how to evolve out of cancer? remember when cancer hasnt take place, good health can significantly lower the risk. But can we promote the health and use this as a treatment for cancer? can we just somehow alter the "environment", say, "tune" the metabolism of an organism to reach a new homeostasis which is not "suitable" for cancer cells to grow? (unfortunately it seems that cancer cells always have better fitness than non-cancer cells... )
王紫薇

1

Threads

11

Posts

61

Credits

Registered member


IP  162.105.160.21

Posted at 2012-5-14 16:47:10 | All floors
   I am very interested in this topic~
   You said “cancer development and spread is fundamentally a micro-evolutionary process” in the Outcomes part of the text.
   I think that knowing what kinds of molecular evolutionary events (such as duplication, horizontal transfer, mutation...) are favored in the process of evolution can help~ Maybe some mechanisms is more powerful than other, which the cancer cell prefer to use. ??

2

Threads

19

Posts

109

Credits

Registered member


IP  162.105.49.113

Posted at 2012-5-15 14:24:43 | All floors
In today's discussion in the Old Chemistry Building, several people, including me, proposed a possible plan for this topic. The first step in our plan is to accelerate the mutation rate in the laboratory condition, and find out which mutants are responsible for drug resistance. At the same time, we shall also estimate the mutation rate, as well as the distribution of these drug-resistant mutants. Next, we shall build models based on the data. Based on what we could learn from the experiments and the models, therapies for drug resistance of cancer could be offered.

As there would be only four days for team challenge, there is no time for performing the experiments we suggested in previous paragraph. However, we can perform in silico experiments in computers. Based on knowledge learned from previous published papers, we can simulated the process based on certain mechanisms.

2

Threads

21

Posts

74

Credits

Moderator


IP  162.105.160.246

Posted at 2012-5-15 16:35:21 | All floors
Reply ChangChang Add Thread


    But how to perform the experiments in computers as we expected?  Which cancer should we focus on and what kind of drugs will be better in our experiment?

1

Threads

28

Posts

146

Credits

Registered member


IP  124.128.37.18

Posted at 2012-5-15 16:59:07 | All floors
Reply YuanYuan Add Thread


    indeed. what do we have to know to start this simulation?
maybe at least some mechanisms as changchang said.
first, we have to know the drugs & their target (some enssential pathways for cell to survive & proliferate, perhaps),
second, the cancer cell.
then it raises the question of ur "cancer network" again no matter if a network is enough or available...
as is known, the gene expression& metabolism of cancer cells are very much different from normal cells,
and what is worse, different from each other among cancer cells.
maybe it is difficult to produce a model even for one kind of cancer,
because there is no reliable rule for a "crazy " cell to behave...
王紫薇

1

Threads

5

Posts

25

Credits

Newbie Member


IP  124.193.192.11

Posted at 2012-5-15 22:51:23 | All floors
I think CADD and the present known drugs or Bioactive compounds may be useful to test this network
You need to log in before you can reply Login | Register now

This forum Credits Rules

Mobile ver.|Darkroom|Complex Systems Discussion Group

GMT+8, 2018-10-15 05:49 , Processed in 0.096675 second(s), 29 queries .

Powered by Discuz! X3 English ver.

© 2001-2013 Comsenz Inc.

Quick Reply Back to top Back to list