Neural network and nonlinear mix effect modeling’s ability to explore both the accuracy of a module informed drug discovery and as a mechanism to streamline drug development process

    Author Name(s)

    Crystal Fotheringham

    Additional Author(s)

    Shaundee Jameson
    Brisa Porter-Garcia
    Daniel Ramirez
    Jessica Tapia

    Faculty Advisor(s)

    Dr. Alona Kryshchenko

    Abstract

    Modeling and simulation also known as model-informed drug discovery and development have become instrumental in streamlining the drug development process. It allows to integrate knowledge from one phase to the next, and from one indication to the next, both in terms of successes and failures. Modeling and simulation integrate both computer-aided mathematical simulation and biological sciences. It uses pre-clinical and clinical data, along with published industry data to elucidate the relationships between drug exposure, drug response, and patient outcomes. The current challenge with pharmaceutical drug modeling is that the process for finding a suitable concentration for a patient can never be precise. The dosage of different pharmaceuticals affects each human body differently. Additionally, the time to find a satisfactory dosage for a patient can be crucial and should be minimized. In this project, we focus on the comparison of two modeling methods such as nonlinear mixed-effect modeling and neural networks which is one of the methods used in Machine Learning to learn the function that relates inputs to outputs. In particular, we would like to compare these methods in terms of accuracy of results and time of convergence.

    Presentation

    Additional Materials

    Leave a Reply

    Your email address will not be published. Required fields are marked *