A Deep-Learning Proteomic-Scale Approach for Drug Design
Computational approaches have accelerated novel therapeutic discovery in recent decades. The Computational Analysis of Novel Drug Opportunities (CANDO) platform for shotgun multitarget therapeutic discovery, repurposing, and design aims to improve their efficacy and safety by employing a holistic ap...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-12-01
|
| Series: | Pharmaceuticals |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8247/14/12/1277 |
| _version_ | 1797501583910502400 |
|---|---|
| author | Brennan Overhoff Zackary Falls William Mangione Ram Samudrala |
| author_facet | Brennan Overhoff Zackary Falls William Mangione Ram Samudrala |
| author_sort | Brennan Overhoff |
| collection | DOAJ |
| description | Computational approaches have accelerated novel therapeutic discovery in recent decades. The Computational Analysis of Novel Drug Opportunities (CANDO) platform for shotgun multitarget therapeutic discovery, repurposing, and design aims to improve their efficacy and safety by employing a holistic approach that computes interaction signatures between every drug/compound and a large library of non-redundant protein structures corresponding to the human proteome fold space. These signatures are compared and analyzed to determine if a given drug/compound is efficacious and safe for a given indication/disease. In this study, we used a deep learning-based autoencoder to first reduce the dimensionality of CANDO-computed drug–proteome interaction signatures. We then employed a reduced conditional variational autoencoder to generate novel drug-like compounds when given a target encoded “objective” signature. Using this approach, we designed compounds to recreate the interaction signatures for twenty approved and experimental drugs and showed that 16/20 designed compounds were predicted to be significantly (<i>p</i>-value ≤ 0.05) more behaviorally similar relative to all corresponding controls, and 20/20 were predicted to be more behaviorally similar relative to a random control. We further observed that redesigns of objectives developed via rational drug design performed significantly better than those derived from natural sources (<i>p</i>-value ≤ 0.05), suggesting that the model learned an abstraction of rational drug design. We also show that the designed compounds are structurally diverse and synthetically feasible when compared to their respective objective drugs despite consistently high predicted behavioral similarity. Finally, we generated new designs that enhanced thirteen drugs/compounds associated with non-small cell lung cancer and anti-aging properties using their predicted proteomic interaction signatures. his study represents a significant step forward in automating holistic therapeutic design with machine learning, enabling the rapid generation of novel, effective, and safe drug leads for any indication. |
| first_indexed | 2024-03-10T03:20:32Z |
| format | Article |
| id | doaj.art-8357c091016d49d0b69aecc92385b37b |
| institution | Directory Open Access Journal |
| issn | 1424-8247 |
| language | English |
| last_indexed | 2024-03-10T03:20:32Z |
| publishDate | 2021-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Pharmaceuticals |
| spelling | doaj.art-8357c091016d49d0b69aecc92385b37b2023-11-23T10:03:28ZengMDPI AGPharmaceuticals1424-82472021-12-011412127710.3390/ph14121277A Deep-Learning Proteomic-Scale Approach for Drug DesignBrennan Overhoff0Zackary Falls1William Mangione2Ram Samudrala3Department of Biomedical Informatics, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY 14203, USADepartment of Biomedical Informatics, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY 14203, USADepartment of Biomedical Informatics, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY 14203, USADepartment of Biomedical Informatics, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY 14203, USAComputational approaches have accelerated novel therapeutic discovery in recent decades. The Computational Analysis of Novel Drug Opportunities (CANDO) platform for shotgun multitarget therapeutic discovery, repurposing, and design aims to improve their efficacy and safety by employing a holistic approach that computes interaction signatures between every drug/compound and a large library of non-redundant protein structures corresponding to the human proteome fold space. These signatures are compared and analyzed to determine if a given drug/compound is efficacious and safe for a given indication/disease. In this study, we used a deep learning-based autoencoder to first reduce the dimensionality of CANDO-computed drug–proteome interaction signatures. We then employed a reduced conditional variational autoencoder to generate novel drug-like compounds when given a target encoded “objective” signature. Using this approach, we designed compounds to recreate the interaction signatures for twenty approved and experimental drugs and showed that 16/20 designed compounds were predicted to be significantly (<i>p</i>-value ≤ 0.05) more behaviorally similar relative to all corresponding controls, and 20/20 were predicted to be more behaviorally similar relative to a random control. We further observed that redesigns of objectives developed via rational drug design performed significantly better than those derived from natural sources (<i>p</i>-value ≤ 0.05), suggesting that the model learned an abstraction of rational drug design. We also show that the designed compounds are structurally diverse and synthetically feasible when compared to their respective objective drugs despite consistently high predicted behavioral similarity. Finally, we generated new designs that enhanced thirteen drugs/compounds associated with non-small cell lung cancer and anti-aging properties using their predicted proteomic interaction signatures. his study represents a significant step forward in automating holistic therapeutic design with machine learning, enabling the rapid generation of novel, effective, and safe drug leads for any indication.https://www.mdpi.com/1424-8247/14/12/1277computational drug designdeep learningmultiscalepolypharmacologyautoencoderdocking |
| spellingShingle | Brennan Overhoff Zackary Falls William Mangione Ram Samudrala A Deep-Learning Proteomic-Scale Approach for Drug Design Pharmaceuticals computational drug design deep learning multiscale polypharmacology autoencoder docking |
| title | A Deep-Learning Proteomic-Scale Approach for Drug Design |
| title_full | A Deep-Learning Proteomic-Scale Approach for Drug Design |
| title_fullStr | A Deep-Learning Proteomic-Scale Approach for Drug Design |
| title_full_unstemmed | A Deep-Learning Proteomic-Scale Approach for Drug Design |
| title_short | A Deep-Learning Proteomic-Scale Approach for Drug Design |
| title_sort | deep learning proteomic scale approach for drug design |
| topic | computational drug design deep learning multiscale polypharmacology autoencoder docking |
| url | https://www.mdpi.com/1424-8247/14/12/1277 |
| work_keys_str_mv | AT brennanoverhoff adeeplearningproteomicscaleapproachfordrugdesign AT zackaryfalls adeeplearningproteomicscaleapproachfordrugdesign AT williammangione adeeplearningproteomicscaleapproachfordrugdesign AT ramsamudrala adeeplearningproteomicscaleapproachfordrugdesign AT brennanoverhoff deeplearningproteomicscaleapproachfordrugdesign AT zackaryfalls deeplearningproteomicscaleapproachfordrugdesign AT williammangione deeplearningproteomicscaleapproachfordrugdesign AT ramsamudrala deeplearningproteomicscaleapproachfordrugdesign |