[2106.04011] JANUS: Parallel Tempered Genetic Algorithm Guided by Deep Neural Networks for Inverse Molecular Design

Inverse molecular design, i.e., designing molecules with specific target properties, can be posed as an optimization problem. High-dimensional optimization tasks in the natural sciences are commonly tackled via population-based metaheuristic optimization algorithms such as evolutionary algorithms. However, expensive property evaluation, which is often required, can limit the widespread use of such approaches as the associated cost can become prohibitive. Herein, we present JANUS, a genetic algorithm that is inspired by parallel tempering. It propagates two populations, one for exploration and another for exploitation, improving optimization by reducing expensive property evaluations. Additionally, JANUS is augmented by a deep neural network that approximates molecular properties via active learning for enhanced sampling of the chemical space. Our method uses the SELFIES molecular representation and the STONED algorithm for the efficient generation of structures, and outperforms other g

5 mentions: @akshat_ai@robpollice@KwhRd100
Date: 2021/06/15 16:47

Referring Tweets

@akshat_ai Check out our new article: t.co/CQKT7IxFKv We were inspired by parallel tempering, SELFIES & S.T.O.N.E.D to create a new algorithm for inverse molecular design 😄 w/ @robpollice, @A_Aspuru_Guzik t.co/3ZkObz5qU2
@robpollice Excited to present JANUS, a genetic algorithm for molecular design that propagates two generations, one explorative and one exploitative, and that can use a neural network for active learning and on-the-fly fitness estimation. t.co/9Km2AgsHzx
@KwhRd100 JANUS:逆分子設計のためのディープニューラルネットワークによって導かれる並列強化遺伝的アルゴリズム 1つは探査用、もう1つは開発用の、2つの母集団を伝播し、高価なプロパティ評価を減らすことで、より高速な最適化を実現。オクタノール-水分配係数で性能実証した。t.co/NtZInIgrSE

Related Entries

Read more [2102.04283] A Systematic Comparison Study on Hyperparameter Optimisation of Graph Neural Networks f...
0 users, 2 mentions 2021/02/11 15:51
Read more [2102.04600] Graph Energy-based Model for Substructure Preserving Molecular Design
0 users, 4 mentions 2021/02/12 15:51
Read more [2106.00927] An Extendible, Graph-Neural-Network-Based Approach for Accurate Force Field Development...
0 users, 3 mentions 2021/06/05 16:46
Read more [2106.04464] Augmenting Molecular Deep Generative Models with Topological Data Analysis Representati...
0 users, 4 mentions 2021/06/11 16:47
Read more [2106.07273] Flexible dual-branched message passing neural network for quantum mechanical property p...
0 users, 2 mentions 2021/06/17 16:48



安定したサイト運営のためにGitHub sponsorを募集しています。