Practical Class about Computational Toxicology for Ph.D. students at Hokkaido University
On July 28-29, 2021, Assistant Professor Dr. Takeda conducted a practical training on Computational Toxicology at the Chemical Hazard Control Expert Course of the Graduate School of Veterinary Medicine, Hokkaido University.
Environmental pollution and wildlife poisoning caused by chemical substances "chemical hazard" is still an international issue as of 2021. In order to deal with this problem, it is essential to properly evaluate the effects of chemical substances on ecosystems, and researchers in the field of veterinary medicine who have insight into a wide range of animal species are expected to contribute. In this course, interdisciplinary faculty members from wildlife science, epidemiology, molecular biology, and computational chemistry give intensive lectures to cultivate expertise in a wide range of fields necessary for chemical hazard control, and train chemical hazard control specialists who can work in industry, government, and academia.
Dr. Takeda, who is also a graduate of this course, taught Computational Toxicology in this intensive course and evaluated the effects of chemical substances on organisms using "in silico" experimental methods with online databases and computer simulations.
Binding of chemical substances to target molecules in vivo, such as proteins, is an important factor directly involved in the expression of drug efficacy and toxicity. On the other hand, it is very costly and time-consuming to obtain these binding abilities experimentally. For example, in the field of drug discovery, it is virtually impossible to experimentally evaluate all the millions of drug candidates, and in the field of environmental toxicology, it is difficult to obtain biological samples from wild animals. In the field of environmental toxicology, it is difficult to obtain biological samples from wild animals.
In contrast, computational methods such as molecular docking and molecular dynamics simulations, which use the three-dimensional structures of chemicals and proteins respectively, can be used to predict the binding ability (≒ toxic effects) of chemicals and proteins in computer simulations. It is becoming possible to predict the binding ability (i.e., toxic effects) of chemicals and proteins by computer simulations using computational methods such as molecular docking and molecular dynamics simulations.
Although animal experiments are currently essential for toxicity testing of chemical substances, these simulation methods are also attracting attention as a "non-invasive" toxicity testing method that does not harm animals. I believe that this method should be actively adopted in veterinary education in order to reduce animal testing in accordance with the so-called "3Rs" principle.
In this practical training, students learned small molecule docking using these methods and homology modeling, which predicts protein 3D structures from wildlife DNA information, by actually using their own computers. In addition, students were able to create wildlife protein structures using "AlphaFold 2", a highly accurate protein structure prediction tool based on deep learning methods created by a Google Group company that was just released in July 2021.
This course has a high international reputation, and in past years many short-term international students from Asian and African countries have come to attend this lecture, but due to the corona disaster, this year's lecture was a hybrid of face-to-face lectures with doctoral students from Hokkaido University and online lectures. However, due to the corona disaster, this year's lecture will be a hybrid of face-to-face lecture with doctoral students of Hokkaido University and online lecture.
The content of this lecture will also be distributed on demand for overseas students.