[KDD 2020] Molecular Inverse-Design Platform for Material Industries - CrossMinds.ai
[KDD 2020] Molecular Inverse-Design Platform for Material Industries
Aug 13, 202017 views
Seiji Takeda
 ,Th,e d,iscovery of new material,s, ,has been the essential force which ,brings ,a, ,discontinuous   improvement   to   industrial   products’ ,performance., ,However, ,the ,extra,vast combinatorial ,design space, ,of  material  structures  exceeds  human  experts’  capability  to ,explore  all,, ,thereby  hampering  material  development. ,In  this ,paper, we present ,a, ,material industry,oriented web platform of ,an ,AI,driven molecular inverse,design system, ,w,hich automatically ,designs  brand  new  molecular  structures  rapidly  and  diversely. ,Di,ff,erent  from  existing  inverse,design ,solutions,,  i,n  this  system, ,the  combination  of  substructure,based  feature  encoding  and ,molecular graph generation algorithms allows a use,r to gain high,speed,  interpretable,  and  customizable  design  process.  Also, ,a ,hierarchical data structure and user,oriented UI provide ,a ,fl,exible ,and intuitive work,fl,ow. ,Th,e system is deployed on IBM’s and our ,client’s cloud servers, ,and has been used by 5 ,partner companies. ,To illustrate, ,actual industrial use cases, we exhibit inverse,design ,of sugar and dye molecules, that were carried out by, ,experimental ,chemists in those client companies. Compared to a general human ,chemist’s standard performance, the mo,lecular design speed was ,accelerated more than 10 times, and greatly increased variety was ,observed  in  the  inverse,designed  molecules  without  loss  of ,chemical realism., ,CCS CONCEPT, ,•, ,Applied  computing, ,→,Physical  science  and  engineering ,→,Chemistry,;, ,•, ,Computing  methodologies, ,→, ,Machine  learning  ,→, ,Machine ,learning ,algorithms ,→,Feature ,selection,;, ,•, ,Mathematics  of  computing, ,→, ,Discrete  mathematics ,→,Graph theory ,→, ,Graph enumeration,;, , ,
KEYWORDS
, ,Cheminformatics;, ,Bioinformatics; ,Feature ,engineering; ,Generative models, ,ACM Reference format:, ,Seiji  Takeda,  Toshiyuki  Hama,  Hsiang,Han  Hsu,  Victoria  A. ,Piunova,  Dmitry  Zubarev,  Daniel  P.  Sanders,  Jed  W.  Pitera, ,Makoto  Kogoh,  Takumi  Hongo,  Yenwei  Cheng,  Wolf  Bocane,tt,, ,Hideaki  Nakashika,  Akihiro  Fujita,  Yuta  Ts,uchiya,  Katsuhiko ,Hino, Kentaro Yano, Shuichi Hirose, Hiroki Toda, Yasumitsu Orii, ,Daiju  Nakano. ,Molecular  Inverse,Design  Platform  for  Material ,Industries. ,Th,e  26th  ACM  SIGKDD  International  Conference  on ,Knowledge  Discovery  &  Data  Mining  (KDD’20,),  USA.  9  p,ages., ,https://doi.org/10.1145/3394486.3403346, , ,1., 
SIGKDD_2020
Recommended