Discovering NDM-1 inhibitors using molecular substructure embeddings representations
NDM-1 (New-Delhi-Metallo-β-lactamase-1) is an enzyme developed by bacteria that is implicated in bacteria resistance to almost all known antibiotics. In this study, we deliver a new, curated NDM-1 bioactivities database, along with a set of unifying rules for managing different activity properties a...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
De Gruyter
2023-07-01
|
| Series: | Journal of Integrative Bioinformatics |
| Subjects: | |
| Online Access: | https://doi.org/10.1515/jib-2022-0050 |
| _version_ | 1797768115168215040 |
|---|---|
| author | Papastergiou Thomas Azé Jérôme Bringay Sandra Louet Maxime Poncelet Pascal Rosales-Hurtado Miyanou Vo-Hoang Yen Licznar-Fajardo Patricia Docquier Jean-Denis Gavara Laurent |
| author_facet | Papastergiou Thomas Azé Jérôme Bringay Sandra Louet Maxime Poncelet Pascal Rosales-Hurtado Miyanou Vo-Hoang Yen Licznar-Fajardo Patricia Docquier Jean-Denis Gavara Laurent |
| author_sort | Papastergiou Thomas |
| collection | DOAJ |
| description | NDM-1 (New-Delhi-Metallo-β-lactamase-1) is an enzyme developed by bacteria that is implicated in bacteria resistance to almost all known antibiotics. In this study, we deliver a new, curated NDM-1 bioactivities database, along with a set of unifying rules for managing different activity properties and inconsistencies. We define the activity classification problem in terms of Multiple Instance Learning, employing embeddings corresponding to molecular substructures and present an ensemble ranking and classification framework, relaying on a k-fold Cross Validation method employing a per fold hyper-parameter optimization procedure, showing promising generalization ability. The MIL paradigm displayed an improvement up to 45.7 %, in terms of Balanced Accuracy, in comparison to the classical Machine Learning paradigm. Moreover, we investigate different compact molecular representations, based on atomic or bi-atomic substructures. Finally, we scanned the Drugbank for strongly active compounds and we present the top-15 ranked compounds. |
| first_indexed | 2024-03-12T20:49:50Z |
| format | Article |
| id | doaj.art-1f931047d7ab413eaff8966dca38a852 |
| institution | Directory Open Access Journal |
| issn | 1613-4516 |
| language | English |
| last_indexed | 2024-03-12T20:49:50Z |
| publishDate | 2023-07-01 |
| publisher | De Gruyter |
| record_format | Article |
| series | Journal of Integrative Bioinformatics |
| spelling | doaj.art-1f931047d7ab413eaff8966dca38a8522023-08-01T05:15:08ZengDe GruyterJournal of Integrative Bioinformatics1613-45162023-07-012026295510.1515/jib-2022-0050Discovering NDM-1 inhibitors using molecular substructure embeddings representationsPapastergiou Thomas0Azé Jérôme1Bringay Sandra2Louet Maxime3Poncelet Pascal4Rosales-Hurtado Miyanou5Vo-Hoang Yen6Licznar-Fajardo Patricia7Docquier Jean-Denis8Gavara Laurent9LIRMM, University of Montpellier, CNRS, 34095Montpellier, FranceLIRMM, University of Montpellier, CNRS, 34095Montpellier, FranceLIRMM, University of Montpellier, CNRS, 34095Montpellier, FranceIBMM, CNRS, University of Montpellier, ENSCM, 34293Montpellier, FranceLIRMM, University of Montpellier, CNRS, 34095Montpellier, FranceIBMM, CNRS, University of Montpellier, ENSCM, 34293Montpellier, FranceIBMM, CNRS, University of Montpellier, ENSCM, 34293Montpellier, FranceHSM, University of Montpellier, CNRS, IRD, CHU, Montpellier, FranceDepartment of Medical Biotechnologies, University of Siena, I-53100Siena, ItalyIBMM, CNRS, University of Montpellier, ENSCM, 34293Montpellier, FranceNDM-1 (New-Delhi-Metallo-β-lactamase-1) is an enzyme developed by bacteria that is implicated in bacteria resistance to almost all known antibiotics. In this study, we deliver a new, curated NDM-1 bioactivities database, along with a set of unifying rules for managing different activity properties and inconsistencies. We define the activity classification problem in terms of Multiple Instance Learning, employing embeddings corresponding to molecular substructures and present an ensemble ranking and classification framework, relaying on a k-fold Cross Validation method employing a per fold hyper-parameter optimization procedure, showing promising generalization ability. The MIL paradigm displayed an improvement up to 45.7 %, in terms of Balanced Accuracy, in comparison to the classical Machine Learning paradigm. Moreover, we investigate different compact molecular representations, based on atomic or bi-atomic substructures. Finally, we scanned the Drugbank for strongly active compounds and we present the top-15 ranked compounds.https://doi.org/10.1515/jib-2022-0050drug discoverymachine learningmultiple instance learningndm-1 inhibitors |
| spellingShingle | Papastergiou Thomas Azé Jérôme Bringay Sandra Louet Maxime Poncelet Pascal Rosales-Hurtado Miyanou Vo-Hoang Yen Licznar-Fajardo Patricia Docquier Jean-Denis Gavara Laurent Discovering NDM-1 inhibitors using molecular substructure embeddings representations Journal of Integrative Bioinformatics drug discovery machine learning multiple instance learning ndm-1 inhibitors |
| title | Discovering NDM-1 inhibitors using molecular substructure embeddings representations |
| title_full | Discovering NDM-1 inhibitors using molecular substructure embeddings representations |
| title_fullStr | Discovering NDM-1 inhibitors using molecular substructure embeddings representations |
| title_full_unstemmed | Discovering NDM-1 inhibitors using molecular substructure embeddings representations |
| title_short | Discovering NDM-1 inhibitors using molecular substructure embeddings representations |
| title_sort | discovering ndm 1 inhibitors using molecular substructure embeddings representations |
| topic | drug discovery machine learning multiple instance learning ndm-1 inhibitors |
| url | https://doi.org/10.1515/jib-2022-0050 |
| work_keys_str_mv | AT papastergiouthomas discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT azejerome discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT bringaysandra discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT louetmaxime discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT ponceletpascal discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT rosaleshurtadomiyanou discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT vohoangyen discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT licznarfajardopatricia discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT docquierjeandenis discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations AT gavaralaurent discoveringndm1inhibitorsusingmolecularsubstructureembeddingsrepresentations |