Home
Professional Experience
Honors and Distinctions
Research
Publications
Machine Learning in Materials Design

 

RESEARCH



Dr. Achenie's work is in several different interdisciplinary fields including process design, molecular modeling, multi-scale modeling, bioinformatics, drug-delivery and uncertainty analysis. He is a pioneer in molecular design, a subset of computer aided product design. This is an advanced simulation model that addresses the systematic design of chemical compounds with desired physical and chemical properties, with the goal of producing computer based "designer" compounds. Molecular design is a valuable tool used to aid bench chemists in narrowing down the range of compounds to synthesize for particular applications. His current research interests are in Agent-Based Multi-Scale Modeling, Molecular Modeling, Biological Modeling and Machine Learning (ML).
    Some Tools Used:
    • Global/Multi-Objective Optimization,
    • Uncertainty Analysis, Interval Analysis, and,
    • High Performance Computing (i.e. Supercomputers).

 

Current/Future Projects

We are currently studying (1) Molecular dynamics simulation of gas separation (CO2/CH4) in organic/inorganic membranes (collaboration with Prof. Ted Oyama, University of Tokyo); (2) the transport of drugs across the blood-brain-barrier (collaboration with Prof. Yong Woo Lee (SBES); (3) oral drug delivery using transport models in conjunction with pharmacokinetic modeling; (4) computational design of materials via machine learning (collaboration with Prof. Hongliang Xin, see "Machine-learning-augmented Chemisorption Model for CO2 Electroreduction Catalyst Screening", http://xingroup.org/publications). In ML we are employing artificial neural networks (ANN), Deep Learning Networks, Self-Organizing Maps (SOM) and Support Vector Machines (SVM); (5) Agent-Based Multi-scale Modeling; (6) Machine learning for diagnosis of autism spectrum disorder (ASD) in children under 6 years old (Collaboration with Prof. Angela Scarpa, Dept. of Psychology). We develop our own computer codes using some combination of Java, C, C++, R, Python and Agent-Based Codes.


Examples of Research we have conducted
  1. Agent-Based Membrane Separation (here the membrane separates CO2 from a mixture of CO2 and CH4; vertical sweep gas is used to collect CO2 and therefore increase selectivity)


    Simulation Video of Demonstration of Membrane Simulation



  2. Agent-Based Transport across Blood-Brain-Barrier (to study mechanism of therapeutic drug molecules)



  3. Agent-Based Chemical Vapor Deposition on a Fixed Substrate (here, ZnS formation and deposition on a Silica Substrate)



  4. Agent-Based Computational Modeling of Hydrogen Storage in a Metal-Organic Framework (here a MOF is used for hydrogen capture; hydrogen is purple, MOF is yellow/blue complex)


    Simulation Video Demonstration of Metal-Organic Framework
  5. .


  6. Advanced Modeling of Oral Drug Delivery and Potential Impact on Personalized Medicine. Here we show modeling of pharmaceutical drug (taken orally) as it travels down the Gastro Intestinal (GI) Tract. The model is used to predict what happens to the drug inside the body.




  7. Use an Engineering Control System (ECS) to model calcium regulation and related pathologies; differentiate Ca-related pathologies with similar pathophysiologies and provide early diagnosis; identify therapeutic targets and propose treatment strategies for Ca-related pathologies.




  8. Interval Global Optimization and Uncertainty Analysis using Interval Analysis (a.k.a. Reliable Computing).