Accelerating Antimicrobial Discovery with Controllable Deep Generative Models and Molecular Dynamics

Das, Payel; Sercu, Tom; Wadhawan, Kahini; Padhi, Inkit; Gehrmann, Sebastian; Cipcigan, Flaviu; Chenthamarakshan, Vijil; Strobelt, Hendrik; Santos, Cicero dos; Chen, Pin-Yu; Yang, Yi Yan; Tan, Jeremy; Hedrick, James; Crain, Jason; Mojsilovic, Aleksandra

doi:10.48550/arxiv.2005.11248

Cited by 4 publications

(10 citation statements)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…a sequence). Depending on the downstream MO tasks, the sequence representation can either be a string of natural amino acids 26,40 , or a string designed for encoding chemicals. In particular, the simplified molecular input line entry specification (SMILES) representation 10 describes the structure of chemical species using short ASCII strings.…”

Section: Representation Of Moleculesmentioning

confidence: 99%

“…The objective of QMO is to search for improved AMP sequences by maximizing similarity while satisfying AMP activity and toxicity predictions (i.e. classified as being AMP and non-toxic based on predictions from pre-trained deep learning models 26 ).…”

Section: Optimization Of Antimicrobial Peptides (Amps) For Improved T...mentioning

confidence: 99%

“…In our experiments, we use QMO to optimize 150 experimentally-verified toxic AMPs collected from public databases 59, 60 by Das et al 26 hemolytic toxicity. Figure 3 shows their cumulative success rate (turning toxic AMPs to non-toxic AMPs) using QMO up to the t-th iteration.…”

Section: Optimization Of Antimicrobial Peptides (Amps) For Improved T...mentioning

confidence: 99%

“…Recent works have exploited latent representation learning and different optimization/sampling techniques for efficient search. Examples include the combined use of variational autoencoder (VAE) and BO [14][15][16][17] , VAE and sampling guided by a predictor 18,19 , VAE and evolutionary algorithms 20 , deep reinforcement learning and/or a generative network [21][22][23][24][25] , and rejection sampling on an autoencoder 26 . On Figure 1.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Optimizing Molecules using Efficient Queries from Property Evaluations

Hoffman,

Chenthamarakshan,

Wadhawan

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Machine learning has shown potential for optimizing existing molecules with more desirable properties, a critical step towards accelerating new chemical discovery. In this work, we propose QMO, a generic query-based molecule optimization framework that exploits latent embeddings from a molecule autoencoder. QMO improves the desired properties of an input molecule based on efficient queries, guided by a set of molecular property predictions and evaluation metrics. We show that QMO outperforms existing methods in the benchmark tasks of optimizing molecules for drug likeliness and solubility under similarity constraints. We also demonstrate significant property improvement using QMO on two new and challenging tasks that are also important in real-world discovery problems: (i) optimizing existing SARS-CoV-2 Main Protease inhibitors toward higher binding affinity; and (ii) improving known antimicrobial peptides towards lower toxicity. Results from QMO show high consistency with external validations, suggesting effective means of facilitating molecule optimization problems with design constraints.

show abstract

Section: Representation Of Moleculesmentioning

confidence: 99%

Section: Optimization Of Antimicrobial Peptides (Amps) For Improved T...mentioning

confidence: 99%

Section: Optimization Of Antimicrobial Peptides (Amps) For Improved T...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Optimizing Molecules using Efficient Queries from Property Evaluations

Hoffman,

Chenthamarakshan,

Wadhawan

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Although many methods used RL to optimize target property, reward-based RL highly provokes overfitting [19]. CogMol [20] was designed to generate drugs for COVID-19, and it used the controlled latent attribute space sampling method [21] instead of RL. Also, Monte Carlo tree search (MCTS) [22,23,24] and a genetic algorithm (GA) [25,26,27] were employed to create a molecular generative model.…”

Section: Introductionmentioning

confidence: 99%

A molecular generative model with genetic algorithm and tree search for cancer samples

Park¹,

Lee²

2021

Preprint

View full text Add to dashboard Cite

Personalized medicine is expected to maximize the intended drug effects and minimize side effects by treating patients based on their genetic profiles. Thus, it is important to generate drugs based on the genetic profiles of diseases, especially in anticancer drug discovery. However, this is challenging because the vast chemical space and variations in cancer properties require a huge time resource to search for proper molecules. Therefore, an efficient and fast search method considering genetic profiles is required for de novo molecular design of anticancer drugs. Here, we propose a faster molecular generative model with genetic algorithm and tree search for cancer samples (FasterGTS). FasterGTS is constructed with a genetic algorithm and a Monte Carlo tree search with three deep neural networks: supervised learning, self-trained, and value networks, and it generates anticancer molecules based on the genetic profiles of a cancer sample. When compared to other methods, FasterGTS generated cancer sample-specific molecules with general chemical properties required for cancer drugs within the limited numbers of samplings. We expect that FasterGTS contributes to the anticancer drug generation

show abstract

Reprogramming Language Models for Molecular Representation Learning

Vinod¹,

Chen²,

Das³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Recent advancements in transfer learning have made it a promising approach for domain adaptation via transfer of learned representations. This is especially when relevant when alternate tasks have limited samples of well-defined and labeled data, which is common in the molecule data domain. This makes transfer learning an ideal approach to solve molecular learning tasks. While Adversarial Reprogramming has proven to be a successful method to repurpose neural networks for alternate tasks, most works consider source and alternate tasks within the same domain. In this work, we propose a new algorithm, Representation Reprogramming via Dictionary Learning (R2DL), for adversarially reprogramming pretrained language models for molecular learning tasks, motivated by leveraging learned representations in massive state of the art language models. The adversarial program learns a linear transformation between a dense source model input space (language data) and a sparse target model input space (e.g., chemical and biological molecule data) using a k-SVD solver to approximate a sparse representation of the encoded data, via dictionary learning. This method achieves the baseline established by state of the art toxicity prediction models trained on domain-specific data using only a standard well-trained text classifier, thereby establishing avenues for domain-agnostic transfer learning for tasks with molecule data.

show abstract

Accelerating Antimicrobial Discovery with Controllable Deep Generative Models and Molecular Dynamics

Cited by 4 publications

References 75 publications

Optimizing Molecules using Efficient Queries from Property Evaluations

Optimizing Molecules using Efficient Queries from Property Evaluations

A molecular generative model with genetic algorithm and tree search for cancer samples

Reprogramming Language Models for Molecular Representation Learning

Contact Info

Product

Resources

About