David Augustin - Homepage

About

Hi, I'm David! I'm a doctoral candidate in Computer Science at the University of Oxford. I'm interested in pharmacological modelling approaches which speed up the drug development process, and improve the treatment of patients in clinical practice. An example of such a modelling approach is pharmacokinetic and pharmacodynamic (PKPD) modelling. PKPD modelling can be used in an academic setting to understand drugs and their interaction with the human body. In clinical practice, PKPD models also have the potential to personalise dosing strategies such that unintended adverse effects of drugs are avoided, and the treatment outcome for each patient is optimised.

Doctoral Candidate in Computer Science

My research evolves around improving the PKPD modelling approach itself, but also involves modelling the PKPD properties of individual drugs, such as the anti-cancer drugs erlotinib and gefitinib.

This fundamentally interdisciplinary research would not be possible without the immensely helpful supervision of Dave Gavaghan, Antje-Christine Walz, Ken Wang, Ben Lambert and Martin Robinson, whose expertise in

Pharmacology
Oncology
Mathematics

Machine Learning
Statistics
Software Engineering

is elementary to the success of my research. In an attempt to make PKPD modelling accessible to the wider community (including pharmacologists and physicians with potentially little programming experience), me and a team of other software engineers are developing the PKPDApp - an open-source web application which allows users to model the PKPD of drugs and predict optimal treatment strategies.

Projects

All
Research
Software

Simulating clinical trials

Simulating clinical trials for model-informed precision dosing: Using warfarin treatment as a use case.

tbd

Filter inference

A scalable nonlinear mixed effects inference approach for snapshot time series data.

PLOS Computational Biology

Heterogeneity affects R

Heterogeneity in the onwards transmission risk between local and imported cases affects practical estimates of the time-dependent reproduction number.

Philosophical Transactions of the Royal Society A

The EU Orphan Drug Regulation 20 Years On

A Quantitative and Legal Analysis.

European Intellectual Property Review

Treatment response prediction

Is model selection unreliable? Find out more in our article.

biorxiv

Electrophysiology modelling

Derivative-based inference for cell and channel electrophysiology models.

IEEE

Chi

Chi is an open-source Python package which is designed for treatment response modelling.

PKPDApp

The PKPDApp is an open-source web application to model and predict the effects of drugs.

Pints

Pints is an open-source Python package for inference of noisy time series.

Resume

My formal education is in Mathematics and Physics, or more broadly speaking - how mathematical concepts can be used to quantify and understand what is happening around us. With the addition of computational skills and software engineering in the more recent years, I am now trying to make some of these concepts applicable to every day problems, such as personalised medication and treatment of diseases.

Education

DPhil in Computer Science

2019 - Present

University of Oxford, Oxford, UK

I am researching the potential of PKPD modelling to inform personalised treatment strategies for patients in clinical practice.

Master of Science in Physics

2016 - 2019

University of Cologne, Cologne, Germany

Specialising in Statistical Physics and Biophysics, I investigated the predictability of evolution in haploids in a dynamic regime where selection is strong and mutations are too rare for mutational interference to occur.

Master of Advanced Studies in Mathematics

2015 - 2016

University of Cambridge, Cambridge, UK

Part III of the Mathematical Tripos in Cambridge is a purely taught course, in which I specialised in High Energy Physics, and learned about Quantum Field Theories, String Theory and General Relativity.

Bachelor of Physics

2012 - 2015

Humboldt University, Berlin, Germany

With a minor in mathematics, I studied the postulated duality between a special class of Quantum Field Theories in Minkowski space time and String Theory in curved Anti-de Sitter space time.

Selected Teaching Experience

Machine Learning

MT 2021

Department of Computer Science, University of Oxford, UK

Advanced Topics in Machine Learning

HT 2021

Department of Computer Science, University of Oxford, UK

Software Engineering & Sustainable Research

MT 2020

Doctoral Training Centre, University of Oxford, UK

Advanced Statistical Physics

2018 - 2019

Institute for Theoretical Physics, University of Cologne, Germany

Quantum Mechanics & Statistical Physics

2018

Institute for Theoretical Physics, University of Cologne, Germany

Classical Mechanics & Electrodynamics

2017 - 2018

Institute for Theoretical Physics, University of Cologne, Germany

Classical Field Theory

2016 - 2017

Institute for Theoretical Physics, University of Cologne, Germany

Contact

I am currently based in Oxford, UK. If you are interested in my research or would like to initiate a collaboration, please feel welcome to get in touch.

Location:

15 Parks Rd, Oxford OX1 3QD

Email:

david.augustin@cs.ox.ac.uk

GitHub:

https://github.com/DavAug