Machine learning emulation of a computational PCR model

One of the first and most famous chemical reactions involving DNA is the PCR reaction. In 2017, we created a computational model of the PCR reaction for use in teaching. With the pandemic, global interest in our PCR simulator has increased considerably, and so did the number of users. This presented a real challenge to the server running the computationally demanding reaction model.

In this talk, I will present a work-in-progress modification to our PCR simulator where supervised machine learning is used to make a fast lightweight emulation of the original differential equation based PCR model. This emulation aims at supporting larger numbers of users without overloading the server.