Wednesday, December 18, 2013

Can (computational) models be trusted?

As a computational modeller I am part of a group of people doing science in a way that was impossible only a few decades ago. A lot of computational modelling combines some of the features of theoretical work (finding out the essential elements of the reality that needs to be captured and creating a computational representation of it) with experimental work (using the computer model as surrogate of the reality to be studied to quickly and exhaustively use it for experiments). Here is an example where Ziv Frankenstein (working with +Alexander Anderson , +Simon Hayward , +Gus Ay and myself) has captured what we thought were some of the essential cellular and microenvironmental players in prostate cancer progression and used that to study how the microenvironment of the tumour can help explain how the tumour evolves.


This week Aeon magazine published this piece by Jon Turney where he asks whether we should trust computational models at all. Computational models allow us to explore very complicated scenarios that would be impossible to study otherwise. The issue the article raises is whether we are beginning to rely too much on these computer models. This is a genuine concern, in some fields (I think +Artem Kaznatcheev might agree with me that social sciences could be one of those) experiments are really hard and the data scarce. This means that the computational models will have to be either too abstract (limiting the detail predictive power of the model) or risk having to make too many assumptions about aspects that are not clear (and thus leading to wrong predictions).

Bottom line: computational models are extraordinarily useful  but depend on having good data and good understanding of what is being studied. We are much better of for having the ability to use them but be careful of detailed predictions when little is known for they are likely to be just guesses.

Wednesday, November 27, 2013

Introducing mathematical oncology to cancer biology and medical students

This Wednesday, as I did last year ago, I am helping with a lecture series entitled Modern Basic Tools of Research at the Moffitt Cancer Center. Last year's was a pleasurable experience and for 2h we talked about different ways to model growth in tumours. First with growth laws describing population change over time (examples of those laws can be found here), then with more mechanistic models where the dynamics of the tumour growth emerge from the way that different cells interact with each other and with their environment.



This year going that far will be more difficult: my 2h have been reduced to 30m. Thus I am trying to change the focus of the lecture and maybe limit my ambitions. Interestingly this could be a blessing in disguise since I might be forced to try to figure out what is the essence of mathematical modelling of cancer and explain that to smart people that do not have a background (or maybe not even an interesting) in mathematical modelling.

And why 30m this year? Because due to time constraints, the same 2h lecture slot will be used to teach about biostatistics, bioinformatics and mathematical modelling. And that is good: since I have been working in this field every biologist and medical professional I met expected me to be a statistician after describing myself as a mathematical oncologist. This is going to be a great opportunity to explain what we have in common but also the ways in which we work differently.

Hoping to not misrepresent excessively what biostastisticians and bioinformaticians do (expect a line saying UPDATE at the end of this post very soon!), their work is incredibly useful to both cancer biologists and doctors since it allows them to figure out what the data says about a specific biological process or clinical trend. It also allows them to know whether that trend or pattern is meaningful or not or whether they have collected enough experimental points or clinical data to make any statement about it.

The best part is that all that extra information usually comes at very little cost to the experimentalists. Mathematical oncologists on the other hand, we tend to be somewhat more difficult to work with since we need to have an understanding of the biological mechanisms underpinning the cancer we are studying. We, for instance, take a look at the diagram after the first paragraph and think: do all these cell interact with each other? if so how? these tumour cells, are they all the same? where do they come from? do they come alone or together? do they usually arrive in the neighbourhood of the other cells in the diagram? if not, do they sit and wait? Maybe some of the questions would be part of the conversation between experimentalists and but many of them seem to arise when mathematical and computational modellers are involved. When implementing these ideas into a computer, ambiguities are not an option. Ideas that might work in your mind or mine come crashing down when subjected to the cold logic of a computer.

It takes time.

The advantage? We can test new hypotheses, generate novel ones, get molecular, clinical, cellular data and integrate it into the model, we can get all those single cell level measuraments and feed them directly into the model, we can take all these population level experiments and figure out what hypothesis explain them better. We can use that to understand the biology of the cancer, to design new biological experiments, to predict better clinical treatments, to hypothethise how new ones would impact patients in the clinic.



If you are an experimentalist you should know that using mathematical model will require you to work in a different way, to ask different questions and to view of your research with different lenses but it is worth it.
By the way, all the figures in the post have been crafted by our very own +Arturo Araujo .

Monday, September 02, 2013

Pint of Science

+Arturo Araujo and myself think that our mission as scientists consists, not only in producing new and interesting discoveries, not only contributing to research a cure to cancer but also in making sure that the society at large understands the nature of what we do and the implications of our work.


Is for that reason that, together with +Angela Rey, we jumped into the opportunity to help +Parmvir Bahia start +Pint of Science US , and offshoot of the UK based Pint of Science, but with a twist: that we will produce regular meetings and podcasts where people in Tampa and beyond will have the chance to interact with our local scientists.

The first of these events will take place tomorrow at The New World Brewery in Ybor City and will feature a friend and colleague at +Moffitt Cancer Center 's IMO: +Jacob Scott. So science, medicine and mathematics will be the ingredients for tomorrow's event at 7pm (9/3/2013). Come if you can, more details at Pint of Science US.




Monday, July 29, 2013

The story of a paper

Together with my good friend and collaborator +Jacob Scott and our new collaborator +Artem Kaznatcheev , we have recently wrote (what I hope you will find) an interesting paper where we explored one of older games (emergence of motility in tumours mostly made of rapidly proliferating cells) and studied what happened when cells grow and reach an edge or a boundary. Normally I would try to recapitulate how did we do our work and what the results are but this time I do not need to bother since my collaborators are so much faster and better than me at this.

For a story of how this work started and how we did this work take a look at +Jacob Scott's latest post.

The work itself has been described in detail by +Artem Kaznatcheev on his own blog in the following post.


Friday, July 26, 2013

Writing grants

Unfortunately, a good deal of the work of a professional biomedical scientist in the US involves writing grants. It might provide a consolation to some to know that almost 100 years ago scientists still had to write grant proposals. I am not sure how many would have looked like this though:


Monday, July 22, 2013

Dr. Arturo Araujo

Congratulations to one of our members, Arturo Araujo on switching his Mr. to Dr. having successfully defended his PhD thesis at University College London last week.
 
Here we can see +Arturo Araujo celebrating with his PhD supervisor +Peter Bentley and his examiners +Alex Fletcher and +Philip Gerlee 

Monday, July 08, 2013

Exploiting ecological principles to better understand cancer progression and treatment

I talked about this before in Google+ but the paper +Alexander Anderson and myself authored at the Royal Society Interface focus is now available for everybody to see. This is a somewhat unconventional type of paper as it is not exactly a research paper, a review paper or an opinion piece but a combination of all those. I must apologise that the article is not in an open access journal but the version in arXiv has all the content (even if it's not as pretty as the journal version). Also, at least for some time, the article can be downloaded from Interface Focus [PDF].

By the way, the visuals of the article, including the following sketch by +Arturo Araujo describing the ecosystem of prostate cancer metastasis to bone, are quite impressive I think.


Sunday, July 07, 2013

More on game theory and the evolution of invasive phenotypes

+Artem Kaznatcheev , a colleague at McGill University, has taken an old paper of ours and decided to do further analysis on it. The emergence of invasiveness is one of the hallmarks of cancer progression and space is likely to play a very important role. For that reason, I used both standard evolutionary game theory and cellular automata to see how much our original results would change if space is explicitly considered. The results can be found here. Now +Artem Kaznatcheev has worked on a Ohtsuki-Nowak transform so that we could still try to understand the role of space without sacrificing the analytical power of game theory. He blogged about it here and the preliminary results look promising. The number of neighbours a tumour cell has impacts the likelihood of motility to emerge in a tumour population. Hope we can learn more about this technique in the next few days now that +Artem Kaznatcheev is in Florida.

Friday, July 05, 2013

Back to blogger

I am back to blogger! Or to be more precise, we are. When I left blogger  a few years ago the idea was to start using a blogging platform that would allow me to share my ideas with likeminded people. That initially was Nature Networks, and eventually Google+. Now we are starting another experiment with blogger, one in which this blog will capture all the news and events related to the CancerEvo group at the Integrated Mathematical Oncology department at the H. Lee Moffitt Cancer Center in Tampa. Hope to see you all very often around these pages!

David and the CancerEvo group.