This page is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Hate to break it to you but a lot of biologists don't think much of systems biology or of Leroy Hood's activities therein (although his previous work on instrumentation is widely respected). He came to speak at our university before a large packed auditorium and his talk was extremely light on details. He also seemed to attribute to systems biology a lot of breakthroughs which were not achieved under the banner of systems biology (so why preferentially fund people who call themselves 'systems biologists'), whilst failing to point out any achievements, past, present or future, where systems biology as a field actively contributed. This was pretty much the unanimous opinion of my research group. Systems biology is nothing but a buzzphrase and it is difficult to believe that someone of Hood's stature does not know this. There are a lot of statements on this page that I find objectionable and I will try to restore a faithful representation of attitudes to systems biology, although as a declared non-believer I understand my contributions may not be welcome. Zargulon 14:32, 14 February 2006 (UTC)
I object to the blanket statements made in the above comments. Just because you have a bone to pick with Lee Hood's seminar, there is no reason to make such sweeping statements about a whole field of science. Systems biology, as described in this article, is a relatively new field and as such, will attract nay-sayers who clearly have an issue with feeling “left out”. This is nonsense. I assure you that declaring yourself a systems biologist is tantamount to drawing a bulls-eye on your chest and asking people to shoot you. It is virtually impossible to get funding to do this stuff and the only reason people are excited about the field is because it offers a hope to finally solve some very important (and extremely difficult) problems in biology.
The salient issue in biology is its extreme complexity. Biologists become reductionist because they simply can’t keep the full complexity of any experimental system in their mind. Thus, they focus on a smaller and smaller piece of the puzzle. Up until now, there simply was no way to keep track of all of the complexity. The advent of computer has provided a way to manage the complexity. The introduction of high-throughput technologies has provided a way to generate the data necessary to describe the complexity in sufficient detail. Mathematical models and statistical approaches offer a way to structure the complexity. If we can handle the complexity, we have a chance to solve big problems.
Can we do full-blown systems biology yet? No. Have we developed the software, the infrastructure or the technologies that we need? Not yet. However, the path we need to take is clear. To say that people have been doing systems biology for decades is absurd, or indicates a lack of appreciation of what is meant by the term. This is going to be an extremely exciting area of research, but it is only in its infancy. It is not a “buzzword”. It is the future of biology. FLeader 22:34, 17 February 2006 (UTC)
Well, I see I have been censored and pigeonholed already.. that was quick. But leaving aside your mischaracterization of my remarks, your arguments are typically ad populum and disingenuous. Saying what "people are interested in", repeating the word "complexity" until it becomes meaningless, are about what I have come to expect. The metaphor of drawing a bullseye is however audacious. For a systems biology grant not to be funded it has to be extremely bad, even for systems biology. For those who draw their excitement primarily from secure access to funds, systems biology is already an "extremely exciting area of research". For your information, it is only in the relatively recent past that people have had the luxury of focussing "on one piece of the puzzle". The greatest biologists have always had sufficient breadth to draw on any sub-field of mathematics or natural science that had the possibility of contributing to their ideas. I certainly don't "think people have been doing systems biology for decades". People have, however been doing biology for decades - millenia even. People have been doing biology with mathematics and statistics for decades if not centuries. People can now do biology, mathematics and statistics with powerful computers. It is still biology.. one name is plenty. Zargulon 01:10, 18 February 2006 (UTC)
There has been a merger on Regulome for a long time. Could someone who knows the subject please either make the merge, or offer an opinion on whether the merge should be done. I don't know the subject well enough. Kcordina Talk 13:37, 22 March 2006 (UTC)
To me, systems biology is the integration of -omic data, with the regulome being just one of the datasources which the systems approach draws from. I disagree with the merge. EdChuong 10:08, 26 March 2006 (UTC)
I added JSim
to the "Software Tools" section April 17, and it was removed without comment by 212.186.70.180 4 days later. Can anyone comment on why this was done? JSim is used primarily for biomedical and physiological modeling, including cardiovascular and pulmonary integrated systems. The MFAX subsystem supports component-based model construction for chemical networks, transport and metabolism. JSim also supports the CellML and (recently) SBML archival model formats. It seems to me it should be mentioned as an available tool. Thanks for your comments.
Erik_Butterworth May 12 19:55 UTC 2006
Given no response to the above comment, I've readded JSim to the "Software Tools"
section. If there is disagreement about this, please discuss.
Erik_Butterworth Jun 12 2006
I redirected a page that indicated it was a synonym of systems biology. Can someone with expertise confirm? [1] Thanks.-- Chaser T 06:36, 11 July 2006 (UTC)
removed lots of external links, this is supposed to be an encyclopedia article not a link farm
Added a link to the wiki systems theory article. It is very nice and concise, and the full content of that article could actually be applied to systems biology..... Moreover, in my opinion the history section should have at least few words for Ludwig von Bertalanffy and Norbert Wiener. Cheers, Igor
I think the POV tag was premature. The article is perfectly stable, there hasn't been any fighting and no-one has complained on the talk page for ages. Zargulon 14:33, 10 October 2006 (UTC)
That's fair enough, the NPOV tag is too strong. However, the lack of citations and the potential inaccuracy of the entry concerns me.
As the disagreement at the top of this discussion page shows, there is a lot of emotion concerning the value and potential of systems biology. There are some very strong views amongst biologists about systems biology. Some think it is only superficially different from other scientific fields, others that it is a fundamental change; some think it will revolutionise biology and medicine, others that it is just a passing trend.
It seems that this wikipedia entry has been used somewhat as a platform for viewpoints on these issues. If it is really necessary, the opinions of recognised scientists could be cited in a clearly defined subsection e.g. "The status of systems biology among scientists". I'm not sure that even this is constructive.
It seems to me this entry could be a lot more encyclopaedic, with improved language, content and citations. robnpov 10 October 2006
Time will tell. Zargulon 15:04, 20 October 2006 (UTC)
Robnpov 15:15, 21 October 2006 (UTC)
Please feel free to rewrite the article according to these sources, they seem to be reliable and I certainly won't revert you. However they seem just as fuzzy as what we have now, and to a certain extent, mutually contradictory (even excluding the current one from the wikipedia article which you feel is suspect). For instance, what are the "functional properties of an organism".. and please don't say "that which is described by systems biology". I certainly understand what the functional properties of a protein are (where the term functional is to distinguish from the properties to do with structure or sequence), but I have yet to be persuaded that this conceptual extrapolation to the organism is anything but doublespeak. Secondly, Kitano's emphasis on dynamics is familiar to me but I don't think it's universal.. the "give rise to life" phrase in the ISB definition touches upon the fact that there are self-described systems biologists who deal with the static or phylogenetic properties of a system. I could go on. The only thread which I feel is universally accepted is that systems biology is an "emerging field". Zargulon 16:48, 21 October 2006 (UTC)
The focus of systems biology is currently on the properties of intra-cellular networks, be they small or large. The goal is to understand systems rather than just components. There are also two parts to sb, systematic, with its emphasis on the organization and interpretation of large-scale high-throughput collection of data sets (essentially bioinformatics on steroids). The second part emphasizes dynamics, which usually involves the dynamic study of small systems and relies on making detailed measurements of time series and kinetic parameters, usually the former. Both approaches to SB are quantitative, with the later (dynamics) being exclusively quantitative. SB is a break from the past because of it's emphasis on systems and quantitative analysis. Traditional molecular biology relied almost exclusively on cartoons as the final repository of knowledge. I also don't consider SB to be an emerging field, for one thing it has been around since the 50s, secondly there is a huge literature on it and thirdly it has a mature set of tools, concepts and methodologies. What makes it seem like an emerging field is that the number of people that did SB in the past wasn't large compared to the total number of molecular biologists so it didn't have a high visibility. What's happened in the last 10 years is that mainstream molecular biologists and researches from other disciplines looking for something different to do have entered the field (Herbert Sauro, 23 Oct, 2006).
I've also taken the libery to edit the current SB content. The latest version of the systems biology (SB) page appears to have some bias, it is too pessimistic (essentially claiming at the end of the page that the goals of SB may never be attained), it is biased towards the data collection part of systems biology (which has been previously called systematic SB, added ref to article section which explains this) and introduces some new terms which I don't think are in common usage in the field (eg Systems Biology measurement and Systems Biology Modeling). The latest version also equated the traditional causal model approach, ie drawing cartoons, to SB, which is certainly not the case. Systems biology is an attempt to break away from this approach and this should be made clear. My edits try to rebalance the article by adding more emphasis to network dynamics and adding a references which backs up this view (Herbert Sauro, 23 Oct, 2006).
I'm not sure who you are Zargulon but the phrase systematic systems biology was coined by some of the leaders in the field (see reference I added) and I for one think it is very apt. It describes precisely the work done by people working in high-throughput molecular biology. If you have *any* constructive criticism, I would be happy to hear it, just saying that you are embarassed to read the page does not really help (Herbert Sauro).
Hello Herbert Sauro!
I welcome your efforts and contributions to the article. (Zargulon, please don't bite the newcomers :) ...)
I agree there should be a greater emphasis on the study of networks and their dynamics. Perhaps we could give a review of the kinds of networks that are commonly studied (e.g. give examples of major collaborations such as YSBN, Hepatosys, Ribosys, etc)?
I think your explanation of the modelling process in systems biology is much better! However, the division into systematic/dynamic systems biology was something I hadn't heard of. I looked at the reference you added -- the WTEC report. They refer to systematic biology and systems biology as distinct: '...there is a distinction between “systems biology” and “systematic biology.”' ( WTEC Final Report, 2005). They seem to have created their own definition of systematic biology, which referred to high-throughput experimentation -- this relates to systems biology -- but I wondered whether they realised there was already a field under that name! Systematic biology is the study of the evolutionary tree. In any case, I didn't find a reference to "systematic systems biology".
I don't see much evidence for widely accepted, named subfields of systems biology. Still I think your explanations are good. So, would you agree to combining the two sections "systematic..." and "dynamic..." -- maybe even incorporate them into the introduction? I think it would be a good step towards a more comprehensive description of systems biology, and it would be less contentious. Robnpov 20:39, 26 October 2006 (UTC)
Just a quick note: I've finally got round to editing the introduction. I've tried to define systems biology both in terms of its goals and in terms of it's approach. Hopefully, in my attempt to be more concrete, I haven't been too narrow in my description: it perhaps needs more examples of the mathematical/numerical techniques for example. It would be good if we could keep the definition broad without being vague. On a curious note, I cited one of the contributions to Alberghina and Westerhoff's book "Systems Biology: Definitions and Perspectives", and noticed that Herbert M. Sauro was one of the contributors to that paper :-) Robnpov 00:21, 27 October 2006 (UTC)
Not wishing to appear to bite anybody, but this systematic systems biology stuff really needs to be removed.. it seems to be based on a misinterpretation of a primary source. It really is in the interest of all you guys who are excited by systems biology to clean this page up. Zargulon 12:04, 31 October 2006 (UTC)
Moved from main article:
Several organizations have prepared extensive reviews of Systems Biology, the definition of the field and its potential socio-economic impact. These includes, example, The European Science Foundation and The Canadian Society for Systems Biology. These documents are free of diciplinary and personal bias, and are considered reliable sources of information on the topic.
The information is of interest, especially for sourcing. It appears to be a response to the discussion on neutrality/citations and as such belongs on the talk pages; in any case it is currently too discussional to be included in the article (e.g. includes weasel term in the last phrase). Robnpov 01:54, 7 November 2006 (UTC)
Most of the information currently in the introduction seems to me to be categorically false, or use ill-defined language, e.g. what is a mechanistic model in this context? Zargulon 10:46, 19 November 2006 (UTC)
A mechanistic model is a model that aims to *explain* how a system works, and not to *describe* its function. This is the big difference between Systems Biology and a phenomenology approach applied to biological systems (i.e. most of the quantitative physiology of the XX century). In order to design a phenomenological model, you consider your system as a black box. You take a set of inputs, you measure the outputs, and then you design a mathematical transformation that produce the set of outputs based on the inputs. There is no mechanist thinking in that approach. Actually, several mathematical descriptions could give the same results. The problem is that such a method is based on a very strong and fragile assumption: that the set of inputs you used to challenge the system adequately represented all the possible sets of inputs. Otherwise, your mathematical transformation will maybe spit-out wrong results. Systems Biology is different. Instead to build a model of level n with data at level n, you build a model of level n with data of level n-1. The output at the level n is therefore an emergent property of the systems, and should (ideally) be valid whatever is the set of input.
My Name is Nicolas Le Novere ( http://www.ebi.ac.uk/~lenov/). I am neurobiologist and bioinfomatician. I have been working in the field of computational systems biology since 1999, when I spent two years working in the group of Dennis Bray, developing models of bacterial chemotaxis (NB Dennis just received a price jointly delivered from the UK Royal society and the french academy of sciences for his modelling work). I am now leader of the group Computational Neurobiology of the European Bioinformatics Institute ( http://www.ebi.ac.uk/compneur/). Besides our work on modelling neuronal signalling, we develop resources and standards for computational modelling. So, Zargulon, what are your credentials, that allow you to be so dismissive of other's work?
I agree with the what you and what you are, but don't you think that the latter modulates the relevance of the former? Anyway, discussion closed. The dismissive stemmed-up from the "categorically false", "ill-defined language", "gibberish". Now, let's go back to more constructive work. Nicolas Le Novere 19:24, 5 December 2006 (UTC)
I am sorry to disagree since I am not a native english speaker, but this is not true. The definition of "mechanistic" is, according to wordnet: "explained in terms of physical forces". "Mechanistic models" are defined exactly as I said. I encourage you to read for instance the following note from Wolfram (I know, he is controversial for his extreme positions on what physics should focus on. But this is not the point here). A mechanistic model is a model that provides an insight on the underlying mechanisms and not an abstract description of the system's behaviour. This is maybe not perfect english, but this is the definition. Now, I did not feel I was dismissive at all of physiology. Quantitative physiology was tremendously useful in the past, and is still incredibly powerful in medicine. It is just a different way of representing biological function. It is based on a very respectable idea, that is we don't care much about the internals of a system if we can isolate it as a module, and describe it correctly in an holistic manner. Similarly in physics, you do not need to use Newtonian mechanic of water molecule in order to describe the tidal waves. However, because of the complexity of biological systems, often we cannot consider them as statistical ensembles. Many people trace Systems Biology to Van Bertalanfy and mention Hodgkin-Huxley as the first "simulation". And they are right. However, one of the first models in Biology predates that. It was a phenomenological model, but the author himself recognised the fact, and advocated for another kind of models, that would not hide the inner details: "One would like to be able to follow this more general process mathematically also. The difficulties are, however, such that one cannot hope to have any very embracing theory of such processes, beyond the statement of the equations. It might be possible, however, to treat a few particular cases in detail with the aid of a digital computer. This method has the advantage that it is not so necessary to make simplifying assumptions as it is when doing a more theoretical type of analysis." This author was Alan Turing talking about the switch from one developmental pattern to another, a non-linear process [AM Turing (1952). The chemical basis of morphogenesis. Phyl Trans Roy Soc Lond B237: 37-72] Nicolas Le Novere 22:19, 5 December 2006 (UTC)
Yes, mechanistic does refer to an enhanced level of understanding when it comes to modelling. Did-you not read the note I mentioned? Just type mechanistic model in google then. Phenomenology indeed does not imply an absence of understanding. When did-I pretend that? A phenomenological model is a model that describes a phenomenon without references to the causes of the phenomenon. Zargulon, I do not invent anything here. This is classical terminology. You may not agree with it, but I can't do anything about it. Please see Phenomenology (science) Nicolas Le Novere 23:36, 5 December 2006 (UTC)
Yes, a mechanistic model does refer to the causes of a phenomenon. And this has nothing to do with physics. We are talking about biology here. Just type "mechanistic models in biology" in Google. And while you're at it, type "mechanistic" or "mechanistic model" in PubMed, and explore. The word mechanistic is not ill-defined in biology. It means precisely "that refers to the underlying mechanisms, or causes". Now, do the same (Google and PubMed) for "phenomenological models", and you will see that a whole community uses this vocable all the time. I do not quite understand your problem with that. And I will stop the discussion here because it goes nowhere. Those concepts are concepts to be used in theoretical biology. Theoretical biologists agreed-upon their definition and use. You can call the sky "green" if you wish. That won't stop most of people to call it "blue". Nicolas Le Novere 17:56, 6 December 2006 (UTC)
Come-on, remove the double-quotes and look a the first of the 965000 hits. Or type it in PubMed and read some of the 294 scientific articles. Nicolas Le Novere 19:27, 6 December 2006 (UTC)
Yes, they do. Of course they do. "Mechanistic models incorporate biological reasoning in the formulation of large-scale ecological systems, metabolic processes, and cell signaling pathways." The problem of those mechanistic models is that you have a very large number of independent parameters, because you represent all the interactions between the entities composing the system. Hence the need for dimensionality reduction. Koonce and Pilla present a possible method to achieve that. You have many others. The problem of course, is then that you loose in expressivity what you gain in parsimony. For instance the new variables do not represent physical entities anymore. For a super-simple example, see the classic cell-cycle models of Tyson 1991 Nicolas Le Novere 23:22, 6 December 2006 (UTC)
I find it hard to believe that time and energy is being spent arguing the meaning of the term "mechanistic model". Everyone in the field of systems biology understands the meaning of this term. It simply means that the relationship between the inputs and outputs of the system are defined in terms of molecules and physical processes that can be quantified (and acted on). One of the primary drivers of the field of systems biology is predictive medicine and to achieve that aim, you need mechanistic models. If you go into a doctor’s office and they do a blood test on you, they will get an answer in terms of specific molecules. With advances in molecular testing, we eventually will be able to measure almost any given molecule. However, to be able to broadly relate this kind of data to disease, we must understand them at a mechanistic level. We currently lack this information in many cases, instead only having correlative information. For example, the PSA blood test is correlated with prostate cancer, but has a high false positive rate because PSA is not part of the mechanism by which prostate cancer grows or spreads. HIV tests are very accurate because they evaluate specific molecules on the virus itself. Clearly, you do not need systems biology approaches to diagnose many diseases, but acurate diagnosis and treatment of many chronic diseases would be vastly improved by using it. Obvious examples are very complex diseases such as type 2 diabetes and cancer. Without a mechanistic model of how these diseases work, it will be extremely difficult to come up with adequate treatments. Phenomenological models are not adequate because they are inherently ambiguous. I also do not understand why we let Zargulon bait us. The fact that he did not know of Herbert Sauro indicates how much he knows about the field of systems biology. Similarly, the debate about the meaning of “mechanistic modeling” is absurd. I know that Wikipedia is open to all contributors, but I would think that anyone who spends so much time arguing about the worthlessness of systems biology would at least try to become familiar with the field. On other discussion boards, we have a term for Zargulon. It is “troll”. FLeader 20:57, 25 December 2006 (UTC)