Options

Introduction: Tyler Backman

edited June 2018 in Chat

Hello! I joined to explore the possibility of applying category theory to Metabolic Flux Analysis (MFA), which is a method of modeling the flux of mass and energy in complex living systems by using carbon labeled with an extra neutron (13C). This is a useful method for both studying human metabolic disease, and for engineering cells to produce carbon neutral biofuels and bioproducts in lieu of petroleum based industrial chemical production. Perhaps if this works well enough we could eventually make petroleum extraction comparatively expensive, and no longer profitable.

Academic background

I am a computational/quantitative synthetic biology postdoc at the Joint BioEnergy Institute (JBEI), Berkeley Lab, and UC Berkeley co-advised by Hector Garcia Martin and Jay Keasling. I did my PhD in computational Bioengineering and drug discovery in the Thomas Girke lab at UCR, and attended John Baez's talks whenever I could, but usually kept quiet in the back. I also have a BS in Physics from Oregon State.

Research interests

I am collaborating with wet lab bioengineers to improve the production of microbial biofuels with Metabolic Flux Analysis (MFA). This is a complex problem, with math methods that were mostly discovered ad hoc, and not connected to any existing theory. I suspect there are powerful methods from both probability theory and category theory which could improve the predictive power and utility of MFA, if the problem can be formulated in the right way, and those methods can be implemented in high performance software that is practical to use. Two recent papers that got me really excited about these possibilities are "Modeling framework for isotopic labeling of heteronuclear moieties" by Borkum et al. and "A compositional framework for reaction networks" by Baez and Pollard. The former includes a Haskell implementation of the theoretical methods they describe for computing metabolic fluxes.

I am also interested in applying ideas from category theory to the bioinformatics software I write. For practical reasons (powerful existing libraries) I mostly use Python and R, which aren't ideal for functional programming, in particular because they have a huge overhead for function calls.

I intend to join the Category Theory course at a later time- after I'm done working through Bartosz Milewski's Categories for Programmers - which I have found very accessible so far, since I have a lot more programming experience than math experience.

Comments

Sign In or Register to comment.