US 11,670,403 B2
Method and apparatus for generating chemical structure using neural network
Kyungdoc Kim, Seoul (KR); Youngchun Kwon, Yongin-si (KR); Misuk Kim, Seoul (KR); Jiho Yoo, Hwaseong-si (KR); and Younsuk Choi, Seongnam-si (KR)
Assigned to SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
Filed by SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
Filed on Feb. 6, 2019, as Appl. No. 16/269,092.
Claims priority of application No. 10-2018-0098373 (KR), filed on Aug. 23, 2018.
Prior Publication US 2020/0066377 A1, Feb. 27, 2020
Int. Cl. G01N 33/48 (2006.01); G01N 33/50 (2006.01); G16C 20/40 (2019.01); G16C 20/70 (2019.01); G06N 3/086 (2023.01)
CPC G16C 20/40 (2019.02) [G06N 3/086 (2013.01); G16C 20/70 (2019.02)] 9 Claims
OG exemplary drawing
 
1. A method of generating a chemical structure by using a neural network apparatus, the method comprising:
inputting a descriptor of a chemical structure to a trained neural network that generates a property value of a property of the chemical structure, the descriptor of the chemical structure representing structural characteristics of the chemical structure and the property of the chemical structure being a characteristic possessed by the chemical structure;
determining an expression region for expressing the property in the descriptor, the expression region comprising a first bit position in the descriptor, wherein the descriptor comprises the expression region and a second bit position outside the expression region; and
generating a new chemical structure by modifying a partial structure in the chemical structure, the partial structure corresponding to the expression region,
wherein the determining the expression region comprises the determining the expression region by applying a layer-wise relevance propagation (LRP) technique to a deep neural network in the trained neural network.