KFU. Card publication. Inverse QSAR: Reversing Descriptor-Driven Prediction Pipeline Using Attention-Based Conditional Variational Autoencoder

- RUS
- 中文
- ES

INVERSE QSAR: REVERSING DES&SHY;CRIPTOR-DRIVEN PREDICTION PIPELINE USING ATTENTION-BASED CONDITIONAL VARIATIONAL AUTOENCODER

Form of presentation

Articles in international journals and collections

Year of publication

2022

��

��

Authors, employees of KFU

Madzhidov Timur Ismailovich, author

Bibliographic description in the original language

Bort W. Inverse QSAR: Reversing Descriptor-Driven Prediction Pipeline Using Attention-Based Conditional Variational Autoencoder / Bort W., Mazitov D., Horvath D., Bonachera F., Lin A., Marcou G., Baskin I., Madzhidov T., Varnek A. // Journal of Chemical Information and Modeling. - 2022.

Annotation

In order to better foramize it, the notorious inverse-QSAR problem (finding structures of given QSAR-predicted properties) is considered in this paper as a two-step process including (i) finding “seed” descriptor vectors corresponding to user-constrained QSAR model output values and (ii) identifying the chemical structures best matching the “seed” vectors. The main development effort here was focused on the latter stage, proposing a new attention-based conditional variational autoencoder neural-network architecture based on recent developments in attention-based methods. The obtained results show that this workflow was capable of generating compounds predicted to display desired activity while being completely novel compared to the training database (ChEMBL). Moreover, the generated compounds show acceptable druglikeness and synthetic accessibility. Both pharmacophore and docking studies were carried out as “orthogonal” in silico validation methods, proving that some of de novo structures are, beyond being predicted active by 2D-QSAR models, clearly able to match binding 3D pharmacophores and bind the protein pocket.

Keywords

Bioinformatics and computational biology,Molecular modeling,Molecules,Pharmacophores,Structure activity relationship

The name of the journal

Journal of Chemical Information and Modeling

URL

https://pubs.acs.org/doi/abs/10.1021/acs.jcim.2c01086

Please use this ID to quote from or refer to the card

https://repository.kpfu.ru/eng/?p_id=281118&p_lang=2

Full metadata record

Field DC	Value	Language
dc.contributor.author	Madzhidov Timur Ismailovich	ru_RU
dc.date.accessioned	2022-01-01T00:00:00Z	ru_RU
dc.date.available	2022-01-01T00:00:00Z	ru_RU
dc.date.issued	2022	ru_RU
dc.identifier.citation	Bort W. Inverse QSAR: Reversing Descriptor-Driven Prediction Pipeline Using Attention-Based Conditional Variational Autoencoder / Bort W., Mazitov D., Horvath D., Bonachera F., Lin A., Marcou G., Baskin I., Madzhidov T., Varnek A. // Journal of Chemical Information and Modeling. - 2022.	ru_RU
dc.identifier.uri	https://repository.kpfu.ru/eng/?p_id=281118&p_lang=2	ru_RU
dc.description.abstract	Journal of Chemical Information and Modeling	ru_RU
dc.description.abstract	In order to better foramize it, the notorious inverse-QSAR problem (finding structures of given QSAR-predicted properties) is considered in this paper as a two-step process including (i) finding “seed” descriptor vectors corresponding to user-constrained QSAR model output values and (ii) identifying the chemical structures best matching the “seed” vectors. The main development effort here was focused on the latter stage, proposing a new attention-based conditional variational autoencoder neural-network architecture based on recent developments in attention-based methods. The obtained results show that this workflow was capable of generating compounds predicted to display desired activity while being completely novel compared to the training database (ChEMBL). Moreover, the generated compounds show acceptable druglikeness and synthetic accessibility. Both pharmacophore and docking studies were carried out as “orthogonal” in silico validation methods, proving that some of de novo structures are, beyond being predicted active by 2D-QSAR models, clearly able to match binding 3D pharmacophores and bind the protein pocket.	ru_RU
dc.language.iso	ru	ru_RU
dc.subject	Bioinformatics and computational biology	ru_RU
dc.subject	Molecular modeling	ru_RU
dc.subject	Molecules	ru_RU
dc.subject	Pharmacophores	ru_RU
dc.subject	Structure activity relationship	ru_RU
dc.title	Inverse QSAR: Reversing Descriptor-Driven Prediction Pipeline Using Attention-Based Conditional Variational Autoencoder	ru_RU
dc.type	Articles in international journals and collections	ru_RU