NRC-VABS: Normalized Reparameterized Conditional Variational Autoencoder with applied beam search in latent space for drug molecule design

Designing an optimal and desired drug molecule structure is a challenging problem. Most of the existing solutions/representations reported in the literature for this problem are complex and time consuming. This is due to larger datasets with longer training periods and long learning dependencies. De...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Expert systems with applications Jg. 240; S. 122396
Hauptverfasser: Bhadwal, Arun Singh, Kumar, Kamal, Kumar, Neeraj
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier Ltd 15.04.2024
Schlagworte:
ISSN:0957-4174, 1873-6793
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Designing an optimal and desired drug molecule structure is a challenging problem. Most of the existing solutions/representations reported in the literature for this problem are complex and time consuming. This is due to larger datasets with longer training periods and long learning dependencies. Deep learning’s generative model can be used to enable chemical modelling to generate molecules without explicit complex molecular rules. However, Deep Learning models (LSTM based VAE) suffer from posterior collapse, larger vocabulary of datasets and sub-optimal latent space searching mechanisms. Motivated by this, we propose a recently researched idea of Normalized Reparameterized conditional Variational Autoencoder with applied beam search in latent space (NRC-VABS). The resulting model with normalized vocabulary, conditionally augmented dataset and revised/reparameterized loss function addresses posterior collapse and constructs continuous and consistent latent space for exploitation by beam search during generation stages. The conditions/properties of desirable molecules are specified through a condition vector and is used while training as well as during generation of drug molecules. Beam search is coined on improved normalized SMILES representation. The idea entails by creating samples with beam search and filtering them depending on their condition and identifying the optimal molecules with desired properties. Normalization also improves the information and reduces complexity in latent space. To address the diversity of the generated molecules, a tunable parameter (D) is also used. Various performance evaluation metrics, such as validity, uniqueness, novelty, accuracy, and Frechet ChemNet Distance are used to evaluate the NRC-VABS on benchmark data sets such as GDB13, MOSES and subset of 250k ZINC molecules. The performance of the NRC-VABS is compared with state-of-the-art peer techniques. NRC-VABS generates molecules at validity range from 92% to 84%, Accuracy 89% to 97% at varied level of diversities (D = 1, D = 2 and D = 3). An application of the proposal in terms interpolation and controlling other (2 of 3) properties by varying one (1 of 3) property at a time. Generating only target molecules with desired properties and maintaining diversity improves novel molecules while greatly reducing time complexity as only novel and desired molecules can be generated. •NRC-VABS improved SMILES notation for reducing complexity in the SMILES.•A solution to posterior collapse is proposed that enhance the NRC-VABS performance.•Latent Space is explored by using beam search algorithm.
AbstractList Designing an optimal and desired drug molecule structure is a challenging problem. Most of the existing solutions/representations reported in the literature for this problem are complex and time consuming. This is due to larger datasets with longer training periods and long learning dependencies. Deep learning’s generative model can be used to enable chemical modelling to generate molecules without explicit complex molecular rules. However, Deep Learning models (LSTM based VAE) suffer from posterior collapse, larger vocabulary of datasets and sub-optimal latent space searching mechanisms. Motivated by this, we propose a recently researched idea of Normalized Reparameterized conditional Variational Autoencoder with applied beam search in latent space (NRC-VABS). The resulting model with normalized vocabulary, conditionally augmented dataset and revised/reparameterized loss function addresses posterior collapse and constructs continuous and consistent latent space for exploitation by beam search during generation stages. The conditions/properties of desirable molecules are specified through a condition vector and is used while training as well as during generation of drug molecules. Beam search is coined on improved normalized SMILES representation. The idea entails by creating samples with beam search and filtering them depending on their condition and identifying the optimal molecules with desired properties. Normalization also improves the information and reduces complexity in latent space. To address the diversity of the generated molecules, a tunable parameter (D) is also used. Various performance evaluation metrics, such as validity, uniqueness, novelty, accuracy, and Frechet ChemNet Distance are used to evaluate the NRC-VABS on benchmark data sets such as GDB13, MOSES and subset of 250k ZINC molecules. The performance of the NRC-VABS is compared with state-of-the-art peer techniques. NRC-VABS generates molecules at validity range from 92% to 84%, Accuracy 89% to 97% at varied level of diversities (D = 1, D = 2 and D = 3). An application of the proposal in terms interpolation and controlling other (2 of 3) properties by varying one (1 of 3) property at a time. Generating only target molecules with desired properties and maintaining diversity improves novel molecules while greatly reducing time complexity as only novel and desired molecules can be generated. •NRC-VABS improved SMILES notation for reducing complexity in the SMILES.•A solution to posterior collapse is proposed that enhance the NRC-VABS performance.•Latent Space is explored by using beam search algorithm.
ArticleNumber 122396
Author Kumar, Kamal
Kumar, Neeraj
Bhadwal, Arun Singh
Author_xml – sequence: 1
  givenname: Arun Singh
  orcidid: 0000-0002-1708-367X
  surname: Bhadwal
  fullname: Bhadwal, Arun Singh
  email: arunsingh.phd19@nituk.ac.in
  organization: University of Petroleum and Energy Studies (UPES), Dehradun, Uttarakhand, 248007, India
– sequence: 2
  givenname: Kamal
  orcidid: 0000-0001-8063-8650
  surname: Kumar
  fullname: Kumar, Kamal
  email: kamalkumar@nituk.ac.in
  organization: National Institute of Technology, Srinagar, Uttarakhand, 246174, India
– sequence: 3
  givenname: Neeraj
  surname: Kumar
  fullname: Kumar, Neeraj
  email: neeraj.kumar@thapar.edu
  organization: Thapar Institute of Engineering and Technology, Patiala, Punjab, 147004, India
BookMark eNp9kMlOwzAURS0EEmX4AVb-gQQPjZ0gNqVikhBIBbq1HPsFXCVxZLsg-AS-mg6sWHR1312cJ91zhPZ73wNCZ5TklFBxvsghfuqcEcZzyhivxB4a0VLyTMiK76MRqQqZjakcH6KjGBeEUEmIHKGfx9k0m0-uni_wow-dbt03WDyDQQfdQYKw6VPfW5ec73WL5zo4_XdPlslDb7yFgD9desd6GFq3AmrQHY6gg3nHrsetTtAnHAdtADc-YBuWb7jzLZhlC9hCdG_9CTpodBvh9C-P0evN9cv0Lnt4ur2fTh4ywwRPWWUrI2VDjWVAjKRCWlY3Y1NW1ghSF1QWgtHamEZYTsaltWVTFiAabmsuOePHqNz-NcHHGKBRxqXNohS0axUlau1ULdTaqVo7VVunK5T9Q4fgOh2-dkOXWwhWoz4cBBWNW1kD6wKYpKx3u_BfKJGVdA
CitedBy_id crossref_primary_10_1007_s11227_024_06250_2
crossref_primary_10_1007_s00521_025_11213_6
crossref_primary_10_3390_ijms25116186
crossref_primary_10_1186_s13321_024_00916_y
Cites_doi 10.1021/acscentsci.7b00572
10.1145/3610533
10.1021/ci990307l
10.1016/S1056-8719(00)00107-6
10.1093/bioinformatics/btac814
10.3389/fphar.2020.565644
10.1186/s13321-018-0286-7
10.2533/chimia.2020.241
10.1186/s13321-023-00696-x
10.1016/j.knosys.2023.110429
10.1093/nar/gkw1074
10.1098/rsif.2017.0387
10.1186/s13321-018-0287-6
10.1016/j.drudis.2018.01.039
10.2533/chimia.2019.1018
10.1145/3388440.3412458
10.1007/s10822-013-9672-4
10.2174/092986709787002817
10.1021/ci049714+
10.1021/acs.jcim.2c00487
10.1093/bib/bbx044
10.1186/s13321-020-00441-8
10.1021/acs.molpharmaceut.5b00982
10.1186/s13321-019-0393-0
10.15252/msb.20156651
10.1038/ncomms13427
10.1021/ci00057a005
10.1021/acs.jcim.8b00234
10.1186/s13321-017-0235-x
10.1162/neco.1989.1.2.270
10.1021/ar500432k
10.1021/acscentsci.7b00512
10.1021/acs.jcim.8b00706
10.1021/acs.jcim.7b00457
ContentType Journal Article
Copyright 2023 Elsevier Ltd
Copyright_xml – notice: 2023 Elsevier Ltd
DBID AAYXX
CITATION
DOI 10.1016/j.eswa.2023.122396
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISSN 1873-6793
ExternalDocumentID 10_1016_j_eswa_2023_122396
S0957417423028981
GroupedDBID --K
--M
.DC
.~1
0R~
13V
1B1
1RT
1~.
1~5
4.4
457
4G.
5GY
5VS
7-5
71M
8P~
9JN
9JO
AAAKF
AABNK
AACTN
AAEDT
AAEDW
AAIAV
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AARIN
AAXUO
AAYFN
ABBOA
ABFNM
ABMAC
ABMVD
ABUCO
ACDAQ
ACGFS
ACHRH
ACNTT
ACRLP
ACZNC
ADBBV
ADEZE
ADTZH
AEBSH
AECPX
AEKER
AENEX
AFKWA
AFTJW
AGHFR
AGUBO
AGUMN
AGYEJ
AHHHB
AHJVU
AHZHX
AIALX
AIEXJ
AIKHN
AITUG
AJOXV
AKRWK
ALEQD
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
AOUOD
APLSM
AXJTR
BJAXD
BKOJK
BLXMC
BNSAS
CS3
DU5
EBS
EFJIC
EO8
EO9
EP2
EP3
F5P
FDB
FIRID
FNPLU
FYGXN
G-Q
GBLVA
GBOLZ
HAMUX
IHE
J1W
JJJVA
KOM
LG9
LY1
LY7
M41
MO0
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
PQQKQ
Q38
RIG
ROL
RPZ
SDF
SDG
SDP
SDS
SES
SEW
SPC
SPCBC
SSB
SSD
SSL
SST
SSV
SSZ
T5K
TN5
~G-
29G
9DU
AAAKG
AAQXK
AATTM
AAXKI
AAYWO
AAYXX
ABJNI
ABKBG
ABUFD
ABWVN
ABXDB
ACLOT
ACNNM
ACRPL
ACVFH
ADCNI
ADJOM
ADMUD
ADNMO
AEIPS
AEUPX
AFJKZ
AFPUW
AGQPQ
AIGII
AIIUN
AKBMS
AKYEP
ANKPU
APXCP
ASPBG
AVWKF
AZFZN
CITATION
EFKBS
EFLBG
EJD
FEDTE
FGOYB
G-2
HLZ
HVGLF
HZ~
R2-
SBC
SET
WUQ
XPP
ZMT
~HD
ID FETCH-LOGICAL-c263t-9d9c77f1cd2e0c7167d2bf4c89dc60b5175621bccf6d3048dd8f85e6f3db37323
ISICitedReferencesCount 5
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001117544700001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0957-4174
IngestDate Tue Nov 18 21:32:28 EST 2025
Sat Nov 29 06:15:34 EST 2025
Tue Jun 18 08:50:57 EDT 2024
IsPeerReviewed true
IsScholarly true
Keywords Posterior collapse
Beam search
Deep Learning
Conditioned Variational Autoencoder
De novo molecule generation
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c263t-9d9c77f1cd2e0c7167d2bf4c89dc60b5175621bccf6d3048dd8f85e6f3db37323
ORCID 0000-0002-1708-367X
0000-0001-8063-8650
ParticipantIDs crossref_citationtrail_10_1016_j_eswa_2023_122396
crossref_primary_10_1016_j_eswa_2023_122396
elsevier_sciencedirect_doi_10_1016_j_eswa_2023_122396
PublicationCentury 2000
PublicationDate 2024-04-15
PublicationDateYYYYMMDD 2024-04-15
PublicationDate_xml – month: 04
  year: 2024
  text: 2024-04-15
  day: 15
PublicationDecade 2020
PublicationTitle Expert systems with applications
PublicationYear 2024
Publisher Elsevier Ltd
Publisher_xml – name: Elsevier Ltd
References Rigoni, Navarin, Sperduti (b38) 2020
Landrum (b20) 2016
Lipinski (b25) 2000; 44
Irwin, Shoichet (b16) 2005; 45
O’Boyle, Dalke (b30) 2018
Yan, C., Wang, S., Yang, J., Xu, T., & Huang, J. (2020). Re-balancing variational autoencoder loss for molecule sequence generation. In
Williams, Zipser (b45) 1989; 1
Bhadwal, Kumar, Kumar (b6) 2023; 268
Angermueller, Pärnamaa, Parts, Stegle (b2) 2016; 12
Arús-Pous, Awale, Probst, Reymond (b3) 2019; 73
Lim, Ryu, Kim, Kim (b24) 2018; 10
Neil, Segler, Guasch, Ahmed, Plumbley, Sellwood (b29) 2018
Gupta, Müller, Huisman, Fuchs, Schneider, Schneider (b14) 2018; 37
Bowman, Vilnis, Vinyals, Dai, Jozefowicz, Bengio (b8) 2015
Lee, Min (b21) 2022; 62
Prasanna, Doerksen (b35) 2009; 16
Olivecrona, Blaschke, Engkvist, Chen (b31) 2017; 9
Arús-Pous, Johansson, Prykhodko, Bjerrum, Tyrchan, Reymond (b4) 2019; 11
Higgins, I., Matthey, L., Pal, A., Burgess, C., Glorot, X., Botvinick, M., et al. (2017). beta-vae: Learning basic visual concepts with a constrained variational framework. In
Mamoshina, Vieira, Putin, Zhavoronkov (b26) 2016; 13
Visini, Arus-Pous, Awale, Reymond (b42) 2017; 57
.
Skalic, Jiménez, Sabbadin, De Fabritiis (b41) 2019; 59
Schoenmaker, Béquignon, Jespers, van Westen (b39) 2023; 15
Gaulton, Hersey, Nowotka, Bento, Chambers, Mendez (b12) 2017; 45
Preuer, Renz, Unterthiner, Hochreiter, Klambauer (b36) 2018; 58
Miotto, Wang, Wang, Jiang, Dudley (b28) 2018; 19
Li, Zhang, Liu (b22) 2018; 10
Arús-Pous, Patronov, Bjerrum, Tyrchan, Reymond, Chen (b5) 2020; 12
Polishchuk, Madzhidov, Varnek (b33) 2013; 27
Krenn, Häse, Nigam, Friederich, Aspuru-Guzik (b18) 2020; 1
Dai, Tian, Dai, Skiena, Song (b11) 2018
Polykovskiy, Zhebrak, Sanchez-Lengeling, Golovanov, Tatanov, Belyaev (b34) 2020; 11
Ozerov, Lezhnina, Izumchenko, Artemov, Medintsev, Vanhaelen (b32) 2016; 7
Kusner, Paige, Hernández-Lobato (b19) 2017
Liao, Xie, Mamitsuka, Zhu (b23) 2023; 39
Wildman, Crippen (b44) 1999; 39
Zhang, Jiang, Cui, Garnett, Chen (b47) 2019; 32
Chen, Engkvist, Wang, Olivecrona, Blaschke (b9) 2018; 23
Segler, Kogej, Tyrchan, Waller (b40) 2018; 4
Alperstein, Cherkasov, Rolfe (b1) 2019
Reymond (b37) 2015; 48
(pp. 1–7).
Kingma, Welling (b17) 2013
Ching, Himmelstein, Beaulieu-Jones, Kalinin, Do, Way (b10) 2018; 15
Weininger (b43) 1988; 28
Gómez-Bombarelli, Wei, Duvenaud, Hernández-Lobato, Sánchez-Lengeling, Sheberla (b13) 2018; 4
Bhadwal, Kumar, Kumar (b7) 2023
Meier, Bühlmann, Arús Pous, Reymond (b27) 2020; 74
10.1016/j.eswa.2023.122396_b46
Neil (10.1016/j.eswa.2023.122396_b29) 2018
Landrum (10.1016/j.eswa.2023.122396_b20) 2016
Arús-Pous (10.1016/j.eswa.2023.122396_b5) 2020; 12
Dai (10.1016/j.eswa.2023.122396_b11) 2018
Bowman (10.1016/j.eswa.2023.122396_b8) 2015
Irwin (10.1016/j.eswa.2023.122396_b16) 2005; 45
Preuer (10.1016/j.eswa.2023.122396_b36) 2018; 58
Liao (10.1016/j.eswa.2023.122396_b23) 2023; 39
Arús-Pous (10.1016/j.eswa.2023.122396_b4) 2019; 11
Wildman (10.1016/j.eswa.2023.122396_b44) 1999; 39
Li (10.1016/j.eswa.2023.122396_b22) 2018; 10
Meier (10.1016/j.eswa.2023.122396_b27) 2020; 74
Reymond (10.1016/j.eswa.2023.122396_b37) 2015; 48
Angermueller (10.1016/j.eswa.2023.122396_b2) 2016; 12
Lee (10.1016/j.eswa.2023.122396_b21) 2022; 62
Schoenmaker (10.1016/j.eswa.2023.122396_b39) 2023; 15
Prasanna (10.1016/j.eswa.2023.122396_b35) 2009; 16
Krenn (10.1016/j.eswa.2023.122396_b18) 2020; 1
Ozerov (10.1016/j.eswa.2023.122396_b32) 2016; 7
Polykovskiy (10.1016/j.eswa.2023.122396_b34) 2020; 11
Weininger (10.1016/j.eswa.2023.122396_b43) 1988; 28
Gómez-Bombarelli (10.1016/j.eswa.2023.122396_b13) 2018; 4
Bhadwal (10.1016/j.eswa.2023.122396_b7) 2023
10.1016/j.eswa.2023.122396_b15
Skalic (10.1016/j.eswa.2023.122396_b41) 2019; 59
Mamoshina (10.1016/j.eswa.2023.122396_b26) 2016; 13
Williams (10.1016/j.eswa.2023.122396_b45) 1989; 1
Kusner (10.1016/j.eswa.2023.122396_b19) 2017
Visini (10.1016/j.eswa.2023.122396_b42) 2017; 57
Segler (10.1016/j.eswa.2023.122396_b40) 2018; 4
Miotto (10.1016/j.eswa.2023.122396_b28) 2018; 19
Kingma (10.1016/j.eswa.2023.122396_b17) 2013
Polishchuk (10.1016/j.eswa.2023.122396_b33) 2013; 27
Arús-Pous (10.1016/j.eswa.2023.122396_b3) 2019; 73
Rigoni (10.1016/j.eswa.2023.122396_b38) 2020
Lim (10.1016/j.eswa.2023.122396_b24) 2018; 10
Ching (10.1016/j.eswa.2023.122396_b10) 2018; 15
Bhadwal (10.1016/j.eswa.2023.122396_b6) 2023; 268
Chen (10.1016/j.eswa.2023.122396_b9) 2018; 23
Gupta (10.1016/j.eswa.2023.122396_b14) 2018; 37
Lipinski (10.1016/j.eswa.2023.122396_b25) 2000; 44
Gaulton (10.1016/j.eswa.2023.122396_b12) 2017; 45
Olivecrona (10.1016/j.eswa.2023.122396_b31) 2017; 9
Zhang (10.1016/j.eswa.2023.122396_b47) 2019; 32
Alperstein (10.1016/j.eswa.2023.122396_b1) 2019
O’Boyle (10.1016/j.eswa.2023.122396_b30) 2018
References_xml – reference: Higgins, I., Matthey, L., Pal, A., Burgess, C., Glorot, X., Botvinick, M., et al. (2017). beta-vae: Learning basic visual concepts with a constrained variational framework. In
– volume: 11
  year: 2020
  ident: b34
  article-title: Molecular sets (MOSES): A benchmarking platform for molecular generation models
  publication-title: Frontiers in Pharmacology
– year: 2019
  ident: b1
  article-title: All smiles variational autoencoder
– volume: 4
  start-page: 120
  year: 2018
  end-page: 131
  ident: b40
  article-title: Generating focused molecule libraries for drug discovery with recurrent neural networks
  publication-title: ACS Central Science
– volume: 9
  start-page: 1
  year: 2017
  end-page: 14
  ident: b31
  article-title: Molecular de-novo design through deep reinforcement learning
  publication-title: Journal of Cheminformatics
– volume: 48
  start-page: 722
  year: 2015
  end-page: 730
  ident: b37
  article-title: The chemical space project
  publication-title: Accounts of Chemical Research
– volume: 59
  start-page: 1205
  year: 2019
  end-page: 1214
  ident: b41
  article-title: Shape-based generative modeling for de novo drug design
  publication-title: Journal of Chemical Information and Modeling
– volume: 39
  start-page: btac814
  year: 2023
  ident: b23
  article-title: Sc2Mol: A scaffold-based two-step molecule generator with variational autoencoder and transformer
  publication-title: Bioinformatics
– volume: 73
  start-page: 1018
  year: 2019
  end-page: 1023
  ident: b3
  article-title: Exploring chemical space with machine learning
  publication-title: CHIMIA International Journal for Chemistry
– volume: 58
  start-page: 1736
  year: 2018
  end-page: 1741
  ident: b36
  article-title: Fréchet ChemNet distance: A metric for generative models for molecules in drug discovery
  publication-title: Journal of Chemical Information and Modeling
– year: 2016
  ident: b20
  article-title: Rdkit: Open-source cheminformatics software
– volume: 62
  start-page: 2943
  year: 2022
  end-page: 2950
  ident: b21
  article-title: MGCVAE: Multi-objective inverse design via molecular graph conditional variational autoencoder
  publication-title: Journal of Chemical Information and Modeling
– volume: 12
  start-page: 1
  year: 2020
  end-page: 18
  ident: b5
  article-title: SMILES-based deep generative scaffold decorator for de-novo drug design
  publication-title: Journal of Cheminformatics
– volume: 44
  start-page: 235
  year: 2000
  end-page: 249
  ident: b25
  article-title: Drug-like properties and the causes of poor solubility and poor permeability
  publication-title: Journal of Pharmacological and Toxicological Methods
– volume: 45
  start-page: 177
  year: 2005
  end-page: 182
  ident: b16
  article-title: ZINC- a free database of commercially available compounds for virtual screening
  publication-title: Journal of Chemical Information and Modeling
– reference: (pp. 1–7).
– volume: 10
  start-page: 1
  year: 2018
  end-page: 24
  ident: b22
  article-title: Multi-objective de novo drug design with conditional graph generative model
  publication-title: Journal of Cheminformatics
– start-page: 1945
  year: 2017
  end-page: 1954
  ident: b19
  article-title: Grammar variational autoencoder
  publication-title: International conference on machine learning
– volume: 268
  year: 2023
  ident: b6
  article-title: GenSMILES: An enhanced validity conscious representation for inverse design of molecules
  publication-title: Knowledge-Based Systems
– volume: 28
  start-page: 31
  year: 1988
  end-page: 36
  ident: b43
  article-title: SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules
  publication-title: Journal of Chemical Information and Computer Sciences
– volume: 10
  start-page: 1
  year: 2018
  end-page: 9
  ident: b24
  article-title: Molecular generative model based on conditional variational autoencoder for de novo molecular design
  publication-title: Journal of Cheminformatics
– volume: 19
  start-page: 1236
  year: 2018
  end-page: 1246
  ident: b28
  article-title: Deep learning for healthcare: review, opportunities and challenges
  publication-title: Briefings in Bioinformatics
– year: 2018
  ident: b29
  article-title: Exploring deep recurrent models with reinforcement learning for molecule design
– volume: 45
  start-page: D945
  year: 2017
  end-page: D954
  ident: b12
  article-title: The ChEMBL database in 2017
  publication-title: Nucleic Acids Research
– volume: 74
  start-page: 241
  year: 2020
  end-page: 246
  ident: b27
  article-title: The Generated Databases (GDBs) as a source of 3d-shaped building blocks for use in medicinal chemistry and drug discovery
  publication-title: Chimia
– volume: 11
  start-page: 1
  year: 2019
  end-page: 13
  ident: b4
  article-title: Randomized SMILES strings improve the quality of molecular generative models
  publication-title: Journal of Cheminformatics
– volume: 39
  start-page: 868
  year: 1999
  end-page: 873
  ident: b44
  article-title: Prediction of physicochemical parameters by atomic contributions
  publication-title: Journal of Chemical Information and Computer Sciences
– reference: Yan, C., Wang, S., Yang, J., Xu, T., & Huang, J. (2020). Re-balancing variational autoencoder loss for molecule sequence generation. In
– year: 2018
  ident: b30
  article-title: DeepSMILES: an adaptation of SMILES for use in machine-learning of chemical structures
– year: 2015
  ident: b8
  article-title: Generating sentences from a continuous space
– volume: 37
  year: 2018
  ident: b14
  article-title: Generative recurrent networks for de novo drug design
  publication-title: Molecular Informatics
– year: 2013
  ident: b17
  article-title: Auto-encoding variational bayes
– start-page: 729
  year: 2020
  end-page: 736
  ident: b38
  article-title: Conditional constrained graph variational autoencoders for molecule design
  publication-title: 2020 IEEE symposium series on computational intelligence
– volume: 15
  year: 2018
  ident: b10
  article-title: Opportunities and obstacles for deep learning in biology and medicine
  publication-title: Journal of the Royal Society Interface
– volume: 32
  year: 2019
  ident: b47
  article-title: D-vae: A variational autoencoder for directed acyclic graphs
  publication-title: Advances in Neural Information Processing Systems
– volume: 57
  start-page: 2707
  year: 2017
  end-page: 2718
  ident: b42
  article-title: Virtual exploration of the ring systems chemical universe
  publication-title: Journal of Chemical Information and Modeling
– volume: 7
  start-page: 13427
  year: 2016
  ident: b32
  article-title: In silico pathway activation network decomposition analysis (iPANDA) as a method for biomarker development
  publication-title: Nature Communications
– volume: 1
  start-page: 270
  year: 1989
  end-page: 280
  ident: b45
  article-title: A learning algorithm for continually running fully recurrent neural networks
  publication-title: Neural Computation
– volume: 23
  start-page: 1241
  year: 2018
  end-page: 1250
  ident: b9
  article-title: The rise of deep learning in drug discovery
  publication-title: Drug Discovery Today
– volume: 16
  start-page: 21
  year: 2009
  end-page: 41
  ident: b35
  article-title: Topological polar surface area: A useful descriptor in 2D-QSAR
  publication-title: Current Medicinal Chemistry
– volume: 27
  start-page: 675
  year: 2013
  end-page: 679
  ident: b33
  article-title: Estimation of the size of drug-like chemical space based on GDB-17 data
  publication-title: Journal of Computer-Aided Molecular Design
– reference: .
– volume: 13
  start-page: 1445
  year: 2016
  end-page: 1454
  ident: b26
  article-title: Applications of deep learning in biomedicine
  publication-title: Molecular Pharmaceutics
– volume: 12
  start-page: 878
  year: 2016
  ident: b2
  article-title: Deep learning for computational biology
  publication-title: Molecular Systems Biology
– year: 2018
  ident: b11
  article-title: Syntax-directed variational autoencoder for structured data
– volume: 4
  start-page: 268
  year: 2018
  end-page: 276
  ident: b13
  article-title: Automatic chemical design using a data-driven continuous representation of molecules
  publication-title: ACS Central Science
– volume: 15
  start-page: 22
  year: 2023
  ident: b39
  article-title: UnCorrupt SMILES: a novel approach to de novo design
  publication-title: Journal of Cheminformatics
– volume: 1
  year: 2020
  ident: b18
  article-title: Self-referencing embedded strings (SELFIES): A 100% robust molecular string representation
  publication-title: Machine Learning: Science and Technology
– year: 2023
  ident: b7
  article-title: GMG-NCDVAE: Guided de novo molecule generation using NLP techniques and constrained diverse variational autoencoder
  publication-title: ACM Transactions on Asian and Low-Resource Language Information Processing
– volume: 4
  start-page: 268
  issue: 2
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b13
  article-title: Automatic chemical design using a data-driven continuous representation of molecules
  publication-title: ACS Central Science
  doi: 10.1021/acscentsci.7b00572
– year: 2023
  ident: 10.1016/j.eswa.2023.122396_b7
  article-title: GMG-NCDVAE: Guided de novo molecule generation using NLP techniques and constrained diverse variational autoencoder
  publication-title: ACM Transactions on Asian and Low-Resource Language Information Processing
  doi: 10.1145/3610533
– volume: 39
  start-page: 868
  issue: 5
  year: 1999
  ident: 10.1016/j.eswa.2023.122396_b44
  article-title: Prediction of physicochemical parameters by atomic contributions
  publication-title: Journal of Chemical Information and Computer Sciences
  doi: 10.1021/ci990307l
– volume: 44
  start-page: 235
  issue: 1
  year: 2000
  ident: 10.1016/j.eswa.2023.122396_b25
  article-title: Drug-like properties and the causes of poor solubility and poor permeability
  publication-title: Journal of Pharmacological and Toxicological Methods
  doi: 10.1016/S1056-8719(00)00107-6
– volume: 39
  start-page: btac814
  issue: 1
  year: 2023
  ident: 10.1016/j.eswa.2023.122396_b23
  article-title: Sc2Mol: A scaffold-based two-step molecule generator with variational autoencoder and transformer
  publication-title: Bioinformatics
  doi: 10.1093/bioinformatics/btac814
– year: 2019
  ident: 10.1016/j.eswa.2023.122396_b1
– volume: 1
  issue: 4
  year: 2020
  ident: 10.1016/j.eswa.2023.122396_b18
  article-title: Self-referencing embedded strings (SELFIES): A 100% robust molecular string representation
  publication-title: Machine Learning: Science and Technology
– volume: 11
  year: 2020
  ident: 10.1016/j.eswa.2023.122396_b34
  article-title: Molecular sets (MOSES): A benchmarking platform for molecular generation models
  publication-title: Frontiers in Pharmacology
  doi: 10.3389/fphar.2020.565644
– volume: 10
  start-page: 1
  issue: 1
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b24
  article-title: Molecular generative model based on conditional variational autoencoder for de novo molecular design
  publication-title: Journal of Cheminformatics
  doi: 10.1186/s13321-018-0286-7
– volume: 74
  start-page: 241
  issue: 4
  year: 2020
  ident: 10.1016/j.eswa.2023.122396_b27
  article-title: The Generated Databases (GDBs) as a source of 3d-shaped building blocks for use in medicinal chemistry and drug discovery
  publication-title: Chimia
  doi: 10.2533/chimia.2020.241
– volume: 15
  start-page: 22
  issue: 1
  year: 2023
  ident: 10.1016/j.eswa.2023.122396_b39
  article-title: UnCorrupt SMILES: a novel approach to de novo design
  publication-title: Journal of Cheminformatics
  doi: 10.1186/s13321-023-00696-x
– volume: 37
  issue: 1–2
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b14
  article-title: Generative recurrent networks for de novo drug design
  publication-title: Molecular Informatics
– volume: 32
  year: 2019
  ident: 10.1016/j.eswa.2023.122396_b47
  article-title: D-vae: A variational autoencoder for directed acyclic graphs
  publication-title: Advances in Neural Information Processing Systems
– volume: 268
  year: 2023
  ident: 10.1016/j.eswa.2023.122396_b6
  article-title: GenSMILES: An enhanced validity conscious representation for inverse design of molecules
  publication-title: Knowledge-Based Systems
  doi: 10.1016/j.knosys.2023.110429
– volume: 45
  start-page: D945
  issue: D1
  year: 2017
  ident: 10.1016/j.eswa.2023.122396_b12
  article-title: The ChEMBL database in 2017
  publication-title: Nucleic Acids Research
  doi: 10.1093/nar/gkw1074
– volume: 15
  issue: 141
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b10
  article-title: Opportunities and obstacles for deep learning in biology and medicine
  publication-title: Journal of the Royal Society Interface
  doi: 10.1098/rsif.2017.0387
– volume: 10
  start-page: 1
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b22
  article-title: Multi-objective de novo drug design with conditional graph generative model
  publication-title: Journal of Cheminformatics
  doi: 10.1186/s13321-018-0287-6
– volume: 23
  start-page: 1241
  issue: 6
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b9
  article-title: The rise of deep learning in drug discovery
  publication-title: Drug Discovery Today
  doi: 10.1016/j.drudis.2018.01.039
– volume: 73
  start-page: 1018
  issue: 12
  year: 2019
  ident: 10.1016/j.eswa.2023.122396_b3
  article-title: Exploring chemical space with machine learning
  publication-title: CHIMIA International Journal for Chemistry
  doi: 10.2533/chimia.2019.1018
– ident: 10.1016/j.eswa.2023.122396_b46
  doi: 10.1145/3388440.3412458
– ident: 10.1016/j.eswa.2023.122396_b15
– volume: 27
  start-page: 675
  year: 2013
  ident: 10.1016/j.eswa.2023.122396_b33
  article-title: Estimation of the size of drug-like chemical space based on GDB-17 data
  publication-title: Journal of Computer-Aided Molecular Design
  doi: 10.1007/s10822-013-9672-4
– year: 2018
  ident: 10.1016/j.eswa.2023.122396_b30
– volume: 16
  start-page: 21
  issue: 1
  year: 2009
  ident: 10.1016/j.eswa.2023.122396_b35
  article-title: Topological polar surface area: A useful descriptor in 2D-QSAR
  publication-title: Current Medicinal Chemistry
  doi: 10.2174/092986709787002817
– volume: 45
  start-page: 177
  issue: 1
  year: 2005
  ident: 10.1016/j.eswa.2023.122396_b16
  article-title: ZINC- a free database of commercially available compounds for virtual screening
  publication-title: Journal of Chemical Information and Modeling
  doi: 10.1021/ci049714+
– volume: 62
  start-page: 2943
  issue: 12
  year: 2022
  ident: 10.1016/j.eswa.2023.122396_b21
  article-title: MGCVAE: Multi-objective inverse design via molecular graph conditional variational autoencoder
  publication-title: Journal of Chemical Information and Modeling
  doi: 10.1021/acs.jcim.2c00487
– volume: 19
  start-page: 1236
  issue: 6
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b28
  article-title: Deep learning for healthcare: review, opportunities and challenges
  publication-title: Briefings in Bioinformatics
  doi: 10.1093/bib/bbx044
– volume: 12
  start-page: 1
  issue: 1
  year: 2020
  ident: 10.1016/j.eswa.2023.122396_b5
  article-title: SMILES-based deep generative scaffold decorator for de-novo drug design
  publication-title: Journal of Cheminformatics
  doi: 10.1186/s13321-020-00441-8
– year: 2018
  ident: 10.1016/j.eswa.2023.122396_b29
– volume: 13
  start-page: 1445
  issue: 5
  year: 2016
  ident: 10.1016/j.eswa.2023.122396_b26
  article-title: Applications of deep learning in biomedicine
  publication-title: Molecular Pharmaceutics
  doi: 10.1021/acs.molpharmaceut.5b00982
– volume: 11
  start-page: 1
  issue: 1
  year: 2019
  ident: 10.1016/j.eswa.2023.122396_b4
  article-title: Randomized SMILES strings improve the quality of molecular generative models
  publication-title: Journal of Cheminformatics
  doi: 10.1186/s13321-019-0393-0
– volume: 12
  start-page: 878
  issue: 7
  year: 2016
  ident: 10.1016/j.eswa.2023.122396_b2
  article-title: Deep learning for computational biology
  publication-title: Molecular Systems Biology
  doi: 10.15252/msb.20156651
– volume: 7
  start-page: 13427
  issue: 1
  year: 2016
  ident: 10.1016/j.eswa.2023.122396_b32
  article-title: In silico pathway activation network decomposition analysis (iPANDA) as a method for biomarker development
  publication-title: Nature Communications
  doi: 10.1038/ncomms13427
– volume: 28
  start-page: 31
  issue: 1
  year: 1988
  ident: 10.1016/j.eswa.2023.122396_b43
  article-title: SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules
  publication-title: Journal of Chemical Information and Computer Sciences
  doi: 10.1021/ci00057a005
– volume: 58
  start-page: 1736
  issue: 9
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b36
  article-title: Fréchet ChemNet distance: A metric for generative models for molecules in drug discovery
  publication-title: Journal of Chemical Information and Modeling
  doi: 10.1021/acs.jcim.8b00234
– year: 2016
  ident: 10.1016/j.eswa.2023.122396_b20
– year: 2015
  ident: 10.1016/j.eswa.2023.122396_b8
– volume: 9
  start-page: 1
  issue: 1
  year: 2017
  ident: 10.1016/j.eswa.2023.122396_b31
  article-title: Molecular de-novo design through deep reinforcement learning
  publication-title: Journal of Cheminformatics
  doi: 10.1186/s13321-017-0235-x
– volume: 1
  start-page: 270
  issue: 2
  year: 1989
  ident: 10.1016/j.eswa.2023.122396_b45
  article-title: A learning algorithm for continually running fully recurrent neural networks
  publication-title: Neural Computation
  doi: 10.1162/neco.1989.1.2.270
– year: 2018
  ident: 10.1016/j.eswa.2023.122396_b11
– volume: 48
  start-page: 722
  issue: 3
  year: 2015
  ident: 10.1016/j.eswa.2023.122396_b37
  article-title: The chemical space project
  publication-title: Accounts of Chemical Research
  doi: 10.1021/ar500432k
– year: 2013
  ident: 10.1016/j.eswa.2023.122396_b17
– start-page: 1945
  year: 2017
  ident: 10.1016/j.eswa.2023.122396_b19
  article-title: Grammar variational autoencoder
– volume: 4
  start-page: 120
  issue: 1
  year: 2018
  ident: 10.1016/j.eswa.2023.122396_b40
  article-title: Generating focused molecule libraries for drug discovery with recurrent neural networks
  publication-title: ACS Central Science
  doi: 10.1021/acscentsci.7b00512
– volume: 59
  start-page: 1205
  issue: 3
  year: 2019
  ident: 10.1016/j.eswa.2023.122396_b41
  article-title: Shape-based generative modeling for de novo drug design
  publication-title: Journal of Chemical Information and Modeling
  doi: 10.1021/acs.jcim.8b00706
– volume: 57
  start-page: 2707
  issue: 11
  year: 2017
  ident: 10.1016/j.eswa.2023.122396_b42
  article-title: Virtual exploration of the ring systems chemical universe
  publication-title: Journal of Chemical Information and Modeling
  doi: 10.1021/acs.jcim.7b00457
– start-page: 729
  year: 2020
  ident: 10.1016/j.eswa.2023.122396_b38
  article-title: Conditional constrained graph variational autoencoders for molecule design
SSID ssj0017007
Score 2.459727
Snippet Designing an optimal and desired drug molecule structure is a challenging problem. Most of the existing solutions/representations reported in the literature...
SourceID crossref
elsevier
SourceType Enrichment Source
Index Database
Publisher
StartPage 122396
SubjectTerms Beam search
Conditioned Variational Autoencoder
De novo molecule generation
Deep Learning
Posterior collapse
Title NRC-VABS: Normalized Reparameterized Conditional Variational Autoencoder with applied beam search in latent space for drug molecule design
URI https://dx.doi.org/10.1016/j.eswa.2023.122396
Volume 240
WOSCitedRecordID wos001117544700001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  customDbUrl:
  eissn: 1873-6793
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0017007
  issn: 0957-4174
  databaseCode: AIEXJ
  dateStart: 19950101
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Li9swEBYh20MvfZduX-jQW3CwJceye0vDlj7AlO52yc3YslwSHGdxnO3Sn9Af0N_bmYzlmGxZ2kIvxhaRZPJ9lmZG82DslQiNzEHLcSKZuY7vGx8-qUA7nmeKQEvtRYaKTag4Dufz6NNg8NPGwlyWqqrCq6vo4r9CDW0ANobO_gXc3aDQAPcAOlwBdrj-EfDx55lzPn1zirp-jBJpufhuMAgRs3yv0Ptl9zxb42E1GQLPQWG2RsHptlljckvMMUGBb1ZONelqtDeSlCCkVs0IFiRNecPzevt1tKJquxiM1XmGLDtvP1M3beroTW_svtEQDQN4Lp-S70C9hdUHNtfOZt15hH9MV2l5rTU2pk6XfUuG2DnAUCwnmddsiM3en4nslMrxPSrlMza0SodKOoGi0op2GReU9unalkDWieXYbL5hnikhxx6IRNFB_u3djn6Kk-FcoJeBIooR_UdCTaJwyI6m70_mH7rzKeVSIL59uTYcizwHD2f6vcjTE2PO7rE7rf7Bp8Sb-2xgqgfsrq3twdul_iH7YWn0mu9JxA9IxHsk4j0S8R6JOALNWxJxJBEnEvFFxYlEfEciDiTiSCJuScSJRI_Yl7cnZ7N3Tlu3w9EikI0T5ZFWqvB0LoyrQSFXucgKX4dRrgM3m4DEGggv07oIcgk7SJ6HRTgxQSHzTCop5GM2rNaVecJ46mVYQUEXWQaiLgyhobPrpaEu_CAV4ph59q9NdJvUHmurlIn1XlwmCEeCcCQExzEbdX0uKKXLjb-eWMSSViglYTMBgt3Q7-k_9nvGbu-_jeds2NRb84Ld0pfNYlO_bHn4C7eCthc
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=NRC-VABS%3A+Normalized+Reparameterized+Conditional+Variational+Autoencoder+with+applied+beam+search+in+latent+space+for+drug+molecule+design&rft.jtitle=Expert+systems+with+applications&rft.au=Bhadwal%2C+Arun+Singh&rft.au=Kumar%2C+Kamal&rft.au=Kumar%2C+Neeraj&rft.date=2024-04-15&rft.pub=Elsevier+Ltd&rft.issn=0957-4174&rft.eissn=1873-6793&rft.volume=240&rft_id=info:doi/10.1016%2Fj.eswa.2023.122396&rft.externalDocID=S0957417423028981
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0957-4174&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0957-4174&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0957-4174&client=summon