High throughput data and multidimensional analysis I

Nowadays it is quite fancy to have high throughput data or some experimental data associated with a mathematical model. These types of work often publish quite well, in high impact factor journals. It requires a strong collaboration between experimentalists and bioinformaticians or theoreticians. One can’t work without the other one, experimentalists often can’t analyze alone the data because they are multidimensional and difficult to interpret without statistical models, and vice versa theoreticians need obviously data, while they lack the knowledge about how to acquire quality data.

What is seldom discussed within the community is how to manage well the collaboration. I would like to talk about some difficulties that are faced in the field. To be concrete I will address the issue of authorship, because it is an essential part of the research activity (sometimes too much!).  The person who performs the majority of the experiments, design the study and interpret the data is usually the first author in classical paper. In the case of high throughput project, the interpretation of the data and even the design of the study are closely related to the bioinformatical analysis. But often the bioinformaticians is still seen as a technical help and thus is offered a position behind the experimentalist in the paper.. The number of people with a mixed background increase more and more, bioinformaticians are now well trained in biology and their work is becoming more than just providing tools for analysis. They really have an impact on the interpretation of the data .Experimental labs often underestimate bioinformatics. I did have the same position before putting a bit of hand in bioinformatics.  But on the other hand the experimentalist also deserve credit because doing experiment is long, complicated and painful!

How can we solve the problem of authorship? First joint authorship is often proposed as a solution. As the number of joint authorship is increasing that’s maybe a good solution but we all know that it is not seen as the same to be the first first author or second first author (I know that it sounds crazy…).  Another option is to split the publication of the methods and the experimental data when possible.  I don’t have better solutions to propose. I think this topic still has to be discussed more broadly within the community of scientists

To end this post, one general comment about the multidimensional analysis. I was surprised how little tools we have to visualize multidimensional experiments. There is urgent need of more input of people working on visualization tools to develop tools for biologists! It is ridiculous to have more tools to analyze facebook network than biological data! And what about the authorship of the people working on visualization?

A rubber arm can tell us a lot of things on ourselves.

A classical experiment in neuroscience and psychology was recently used in a creative way to question the immune theory of self and non-self (1).

The rubber illusion arm experiment was original perform by Botvinick and Cohen in 1998 (2). They have shown that people can assimilate a rubber arm to their own arm. How? There is different ways to perform this experiment but generally synchronous tapping on a person’s hidden real arm and an aligned visible rubber arm placed in front of them results in a feeling of ownership of the fake arm. Some different setups of the experiment and results are described here (3).  This experiment is based on an optical illusion, it is crucial that the rubber look like a real one and it has to be align with the real arm. You can try to reproduce the experiment yourself; it is quite funny!

What is the link with immunology? This experiment really plays with the definition of self, classically viewed as our body. Our brain seems to be able to take a rubber arm for our real arm. In immunology, the classical theory is that the immune system reacts to elements that are not part of our self, so called non-self. This theory was initiated by Burnet and has been criticized recently. I already talk a lot about that in my blog.

Barnsley at al wonder in the situation where a rubber arm become our real arm what happens to the real arm. Is the real arm rejected by the immune system? They report an increase in histamine activity in the real arm compare to the other arm. The study is still a pilot and results have to be confirmed.  Moreover the link with an immune rejection still has to be demonstrated formally.

Still it is a very new way to think about the self and non-self theory. It may confirm the theory but adding new elements to the molecular mechanisms already proposed. It questions the role of the brain perception in the definition of the self and it links with immune response. It is going against the danger theory of Matzinger. How can having two arms will be a danger for your body? I can’t wait to see what will come out of these type of study.

References:

1 N. Barnsley1,2, J.H. McAuley1,2, R. Mohan1,2, A. Dey1,2, P. Thomas2,3,and G.L. Moseley1,4 . (2011). The rubber hand illusion increases histamine reactivity in the real arm. Current Biology, vol 21 No23

2. Botvinick, M., and Cohen, J., (1998). Rubber hands ‘feel’ touch that eyes see. Nature 391, 756–756.

3. Makin, T., Holmes, N., and Ehrsson, H. (2008). On the other hand: Dummy hands and peripersonal space. Behav. Brain Res. 191, 1–10.

Ed Cohen seminar on the 13rd of march 2012

The next Philosophy & Immunology, that I can’t organized anymore from the Netherlands, will  received

Ed Cohen, Université de Rutgers, USA (http://womens-studies.rutgers.edu/faculty/core-faculty/122-ed-cohen) :

on the 13rd of mars 2012, at noon.

It sounds to be challenging! If you are in Paris at that time, you should definitly go!

“If Immunity Doesn’t Exist Is It Still Real? or, A Vital Paradox”

Abstract :

At the end of the 19th century biology and medicine recruited an ancient legal and political concept, “immunity,” to describe what they newly recognized as an essential physiological function of multi-cellural organisms, “defense,” another concept which not coincidentally also had it’s own long legal and political history. Indeed, when Eli Metchnikoff introduced the rubric “immunity-as-host-defense” into bioscience, he actually combined two contradictory terms (i.e., juridically speaking where immunity exists there is no need of (legal) defense, and conversely where a defense is needed there is no immunity) in order to encompass what he imagined as the complex organism’s response to microbial “invasion.” While this formulation proved essential in making sense of the “germ theory of disease” promoted by Koch and Pasteur among others, it also imported a number of philosophical and political assumptions into biomedicine that may not have been warranted by the empirical observations themselves. In particular, Metchnikoff’s hybrid immunity-as-host-defense invoked a specific set of axioms derived from liberal political philosophy in order to affirm that multi-cellular organisms actively conserve their own integrity by defending what is “properly” their own–or even, what is their own “property”–at the cellular level. (Paul Ehrlich, who shared the 1905 Nobel prize with Metchnikoff, would then extend this precept to the molecular level via the concept of “specificity” a decade or so later.)

Fifty years after Metchnikoff first posited immunity as host defense, MacFarland Burnet transformed the field of immunology into the “science of self/non-self discrimination.” In so doing he rendered the notion of the defended organism theoretically plausible by asserting “the self” as the both ontological and epistemological ground of the organism. While the experimental programs that resulted from immunology’s embrace of S/NS discrimination have produced numerous significant insights both about how organisms of different scales co-exist in shared spaces and about how multicellular organisms persist in time, Burnet’s assertions (like Metchnikoff’s) incorporated a number of assumptions that may not have derived either from his experiments themselves or from scientific discourse more generally, but may instead have ensued from the political and philosophical frameworks within which he worked. 

Recently a number of exciting new immunological theories have challenged Burnet’s S/NS doctrine in order to reconsider several persistent impasses in immunological thinking: e.g., autoimmunity, host-versus-graft disease, cancers, pregnancy, and commensal organisms. However, despite the exciting possibilities that these theories present for immunological thought and practice, they nevertheless seem to share a conceptual opacity that has endured since Metchnikoff and Erlich first construed immunity as a robust scientific concept: i.e., they fail to appreciate that immunity is always already a paradoxical term. In fact, in its juridical and political valence, immunity was created precisely in order to rectify the constitutive tensions between the formal structure of the law and the empirical messiness of political life. In other words, immunity was legally invented in order to maintain the fiction that the law’s jurisdiction is universal by creating legal exceptions that affirm the law’s universality even while acknowledging–and moreover finessing–the particular political exigencies that the law inevitably must exempt. 

Unfortunately when biology and medicine incorporate immunity as an organizing principle for their activities they by and large ignore the concept’s foundational status as a paradox. Instead they usually seek to contain the effects of this foundational paradox by construing it in terms of mutually exclusive oppositions (inside/outside, self/not-self, dangerous/not-dangerous, continuous/discontinuous, etc.) thereby displacing the troubling perturbations to which immunity actually gives rise. In this talk, I will try to suggest that whatever it does at the cellular, molecular, or even quantuum-mechanical levels, biological “immunity” as such may not exist insofar as it always “is” paradoxical–which does by any means imply that it is not real. Rather it may be worth considering whether what we have come to call biological immunity in fact specifies the paradox that living organisms must be insofar as they must be simultaneously open and bounded in order to live at all. Following Francisco Varela’s reflections on the “intriguing paradoxicality proper to an autonomous identity,” which he attributes to the fact that “the living system must distinguish itself from the environment while at the same time maintaining its coupling,” this talk will wonder if it could be interesting to consider that biological immunity might “be” precisely that which makes this vital paradox both possible and liveable.

Santa Fe II

I have been once again in Santa Fe for the second conference on system immunology. I will not talk a lot about the content of the conference; you can have a look at the program online. It is more the occasion of doing a little bit of sociology of immunology.

My abstract was selected for oral presentation, which scared me a lot. Even more when I realized that I was the youngest speaker. The other speakers were all senior postdoc or faculty member. The experimentalist speakers were even more senior. It indicates us that the field is still young. Labs don’t spend money to send PhD student  to this type of conference and the field is still more for well-known experimentalists. But things are changing!

Appart Americans, the Duch and English researcher are the more active, with the lab of Robin Callard and rob de Boer. French people were also well represented (5 people). It was not the case 2 years ago and it is great news!

I have noticed a clear gap between two types of models. The first type of models tries to stick to the data and therefore is very simple. The number of parameters that you can estimate limits you. The second type of models is far away from the data. It could still be very interesting but the speaker fails to make the link with the current experimental research. Therefore they have little impact on experimentalists and increase the gap between them and modeling. I was shocked by a multi-scale modeling with 1000 differential equations. This type of model can tell everything you want. You have no way to estimate the parameter and it can show that my grand mother is skiing right now.

To finish I would like to say that I really enjoy this conference. It is a very important one in the field. I also like the discussion sessions that were scheduled.  It is the first time that I attended a conference where we had discuss epistemological issues, such as the definition of the world memory apply to the cells.  No clear conclusions were drawn but the more important was to discuss!

http://www.cvent.com/events/systems-approaches-in-immunology/event-summary-8948c40dcb0d4c2f8736de1a28aacd64.aspx

Let’s start to build a model

It is a long time since I wrote the last post, I have been busy by learning through practice modeling and bioinformatics.  I am happy to start here a series of article on my new experience of modeling.

When I have started in September 2009 at college de France, I had a project with a master student about modeling the differentiation of T helper cells. I remember that I was always telling him that his attempts of model were always too simple, compared to the complexity of this process. I was aware that having a model with a lot of parameters was not a good idea, because you can’t estimate many parameters on the data. But still I was frustrated by the simplicity of the model. I did a lot of work to put altogether the knowledge in the field in a schematic representation and started to simplify it. I found a level of simplification that satisfied me but it was still too complicated and the model failed to be developed.

After 8 months of modeling in Rob de Boer’s lab, I just had a look of what I have been doing. My surprise was big when I realized that I have been doing the complete opposite. In fact I have started with a very simple model that we know it is not realistic, but this type of model has more value to me now. Not because I did it but more because I could explore and use this model to understand why it was not realistic.  For example we know that immune cells in response to a pathogen don’t have the same growth rate over time.  Still being able to describe the data with a constant growth rate makes me understand that there is a mean growth rate that is the sum of all individual growth rates over time.  It was a necessary step before applying a changing growth rate over time.

There is a lot of knowledge to gain from very simple models, even if they are not realistic. These simple models are easy to manipulate. When you fully understand the simple model, you can add step by step some complexity. Moreover you completely control the complexity introduces. It is very useful for biologist because it is the opposite from what we experience with experiments. There the complexity is uncontrolled.

My final world will go to education. I see a clear need of educating biologists to understand the usefulness of starting with simple model.

Natural infection in mouse house: an opportunity or a problem ?

Today in my lab, an infection (Spironucleus muris) has been diagnosed on the mice of our animal facility. The bad news is that infection will ruin all the results from immunological studies because the infection will interfere with the immunological process we wish to study.

Having an infection in animal facilities is quite well-known phenomena for a “mouse” immunologist (an immunologist who performs experiment with mice) but is new for me.  Before I used to work on human cells and organs.

Rather than being desperate, it brings me some fresh air. It is an occasion to study a natural infection. For those of you who are not familiar with immunology, if you were looking at our protocol you could imagine chicken ovalbumine is the worst cause of infection on Earth. Ovalbumine is indeed used as a model in immunology because mice don’t eat chicken and it is supposed that they have never been exposed to chicken protein. As a result, you can control better your experimental setup. Moreover ovalbumine is easy to produce in large quantity as it is the main component of eggs.

A natural infection could be seen as a real opportunity. First, it allows to study the ecology of an infection in mice in a lab. It could be a model as good as ovalbumine. I was really disappointed not to find more articles on that topic on pubmed. Immunologists could try to develop vaccines or treatments for mice.. Moreover we could think that having a patent  and selling a treatment or vaccine could finance your research as the problem is not isolated.