| Name: | Sycamore Herlihy |
|---|---|
| Email: | slherlihy@wpi.edu |
| Institution: | Worcester Polytechnic Institute |
| Mentor: | Dr. Subhajyoti De |
| Project: | Genomic Data-Guided Mathematical Modeling of Cancer |
My main focus this summer is on furthering the development of a model for simulating the evolution of tumors. Specifically, the model focuses on the concept of "intratumoral heterogeneity" - the common phenomenon in which human tumors contain cells with widely varying mutations despite the fact that almost all cancers start from a single cell. This model utilizes stochastic processes to simulate various biological processes that occur during tumor growth.
This week began with an orientation to the DIMACS program. It was nice getting to know some other participants, as well as learn more about what our workday will look like this summer. On 6/2, I had my first meeting with my mentor to go over what I will be working on. Since I didn't understand a lot of the biological concepts, I decided to read some papers I found about intratumoral heterogeneity in order to prepare for my introductory presentation on 6/6. The presentation went great! After the presentation, I headed to the Cancer Institute of New Jersey to meet in-person with my mentor and his research group of grad students. They talked about their recent findings, which I have found very fun. For the most part what I have worked on so far is creating an interface for the tumor growth simulation my mentor had previously developed. It is coming together slowly, but surely.
This week, I have been continuing the interface work from last week. UI is something that I find very enjoyable, so I am having fun creating the interface. I am encountering a lot of bugs.
With the help of my mentor, I finally figured out the last bug in my code that was preventing me from being able to fully run the simulation last Thursday. Now, I can move onto testing to make sure that my interface is set up correctly. Additionally, I have been reading the simulation code more in depth; while before I was just developing the interface, I now want to understand how exactly the code that my mentor gave me works. I corrected a couple errors in the code, such as giving the simulation more random chance (where before it could be slightly biased toward one type of mutation). Finally, in my meeting with my mentor this Thursday we spoke about the possibility of publishing this program with a paper. I am very excited about this, since it would mark my first publication.
This week has been focused on adding more features to the simulation. Whereas before only one model of evolution was available for simulation, I am now adding more va variety in what the user can choose. The program is coming together very nicely. With the addition of the new models, I have been playing around with the parameters of the simulation in order to see how different variables affect the growth. It is very interesting to see that a lot of it is based on purely random chance. Even running the simulation several times with the exact same parameters can yield vastly different results; much like how cancer really develops, as no patient is the same.
This week, I have been continuing my work from the previous week in adding more versions of the simulation. I have also made some efficiency changes. The program is now multi-threaded, which in addition to making it faster also allows the addition of the progress bar for UX, since the simulation can take some time as the number of tumor cells accumulate exponentially. Additionally, I have added another relevant graph to the simulation output, and made some slight changes in regards to the aesthetics of the interface.
On 7/7 I met with my mentor to show some figures for modelling specific models of tumor growth. The figures looked good, however, my laptop can only handle modelling up to 10000 cells. This is an issue because most clinically detectable tumors have at least 10 million cells. I am either going to find a way to run it on a desktop computer, or adapt my code to work with MacOS so that my mentor could run it on his computer.
I decided to start on adapting my code to work on MacOS, because as of now there are some utilities that only work with Windows 10. Towards the end of the week I began working on my final presentation; I am presenting next Friday 7/22. It feels unreal that we are already approaching the end of the program!
This week I continued working on adapting my code for Mac. Additionally, I have never used Github before, so I decided to create a private repository for my code in order to figure out the basics.
My presentation on Friday went great! I am now starting to work on my paper. It's a bit difficult writing the sections that are more about the cancer biology, since that is not in my background, but I am getting through it alright.
The code for MacOS is finally finished! I sent it to my mentor and he was able to simulate up to 250,000 cells in about 1 hour, since his desktop computer runs much faster than my laptop. I am mostly done with my paper by now, just putting some final touches and including some additional figures.
Thank you to:
Presentations
Here is the REU website: