- 🔭 I’m currently working on: OpenFF benchmarking for Sulfonamides and ABFE calculations for RIPK2 .
- 🌱 I’m currently learning: Dash
- 🤔 I’m looking for help with: Designing and deploying Dash apps
- 💬 Ask me about: Cheminformatics, Molecular Dynamics
- 📫 How to reach me: [email protected]
- 😄 Pronouns: they/them
- Package: Simulations - a package to run MD simulations using OpenMM. In particular this will run simulations after docking with AutoDockVina (or any simular outputs) and run a fingerprint analysis.
- Project: Deep online learning of molecular kinetics - WIP. Online (i.e., 1 epoch) learning of the Koopman operator for molecular kinetics using VAMPnets.
- Package: Celerity - WIP - package containing the 'online' deep-learning estimators for molecular kinetics
- Presentation: ScotChem Presentation
- Paper: Markov state model sensitivity and hyperparameter optimisation - Investigating the sensitivity of Markov state models to pipeline hyperparameters and suitability for Bayesian optimisation.
- Published Paper
- Manuscript
- Results analysis
- Data generation
- Deprecated - results analysis. Old version of results analysis. Has some useful code but is not current.
- PhD: Markov models of Biomolecular systems
- Thesis - my thesis.
- Chapter 3 - MSM analysis of water diffusing in a sucrose matrix. Published paper here
- Chapter 4 - Using Bayesian optimistaion to optimise MSMs of alanine dipeptide.
- Chapter 5 - Choosing Hidden Markov Model order with the Integrated Complete Likelihood Criterion. NB - the containing repo contains lots of old/irrelevant analysis.
- Chapter 6 Pocket and ligand dynamics of AADH
- Analysis - Optimising MSM and HMMs of the enzyme AADH.
- Production MD - Create the production AADH simulations using Amber MD.
- System prep - Preparing the system using CHARMM.
- Initial analysis - Some initial cleaning and analysis. Superceded by
Analysisabove.
- Optimising FS-peptide - This didn't make it into the thesis but was useful preparatory work for chapter 4.
- Mal-L - this work formed part of the PhD thesis of Michael Connolly and is available here - see sections 2.2.3, and chapter 5.
- Package: InvSpec - A package that generates stochastic matrices with given eignevalue spectrum (InvSpec = Invert Spectrum).
- Package: Adaptive - A package the tests coverage times for graph exploration using various adaptive sampling strategies.
- Project: Multitask learning for AMR - The JGI funded project MTL for AMR. See here for more details.
- Osprey - Hyperparameter tuning package. Project is no longer in active development.
- PyEMMA - Markov state models. Project is no longer in active development.
- Paper: Sonifiying Stochastic Landscapes - Sonification of hidden Markov models.
- Sonification - The code to produce the parameters to be sonified.
- Performance: Metastable Impressions - The JGI funded project Metastable Impressions a multimedia data visualisation and sonification project. See here for more details.
- Package: HMM to OSC - A package to take a PyEmma hidden Markov model and turn the features into OSC messages for sonification with, e.g., Max MSP.
- Art of MD - Some training materials for the CCPBioSim funded Art of MD course. Using deep learning style transfer for illustrating biomolecules. Some examples of output from the course can be found here (please request access).
- Deep Art - A JGI skill sharing course on art generated by Deep learning.
- Maximum Caliber Explainer - Some notes on the theory behind Maximum Caliber methods for MSMs.
- robert-arbon - My hugo website: robertarbon.com
- Old website - My old static Jekyll Website.
- Wordle Solver - WIP. My attempt at a maximum entropy solver for Worldle (currently has a bug).
- Perlin Noise Minimizer - A script that minimizes and objective function by performing simulated annealing with perlin noise.
- Sonification of 2FD - abandoned sonification project.
- Integer Mapping - some scratch notes on the behaviour of integer warping in Numpy.
- AADH reference reaction - some old AADH system prep using CHARMM