27 July 2018 Vol 361, Issue 6400
Special Issue – Frontiers in computation
Introduction to special issue
By Michael Funk, Colin Norman, Keith T. Smith, Jelena Stajic, Jake Yeston
Science27 Jul 2018 : 342-343
Scientists learn about the world by observing, manipulating, measuring, and abstracting. To make sure that they truly understand their system, and to gain insight beyond what experimental data can provide, many also turn to building mathematical models. Some models are based directly on fundamental physical laws, but most rely on approximations. The computational costs vary widely—from exactly solvable models to those that require all the computer power you can get. This power has become increasingly accessible; throw in the recent forays of artificial intelligence into scientific computing and the tantalizing promise of quantum computers, and it’s clear that exciting times are ahead.
In this special issue, we highlight several areas of scientific computing that have seen recent progress. A feature article by Science reporter Paul Voosen focuses on an ambitious and controversial project, backed by high-tech philanthropists, to leverage breakthroughs in artificial intelligence, satellite imaging, and high-resolution simulations to create a new climate model. Kent and Kotliar reflect on two complementary approaches to predicting the properties of materials in which electrons are inextricably correlated with one another. Bottaro and Lindorff-Larsen report on the simulation of biological molecules, where recent progress has been driven by advances in theory and methods and tighter integration of experimental constraints. Sanchez-Lengeling and Aspuru-Guzik review approaches that use machine learning to design chemical structures that bear a desired characteristic. Finally, Brügmann describes how to simulate the complex equations that govern four-dimensional spacetime in Einstein’s theory of general relativity, focusing on the example of gravitational waves produced by merging black holes.