Enhancing reinforcement learning for de novo molecular design applying self-attention mechanisms.

The drug discovery process can be significantly improved by applying deep reinforcement learning (RL) methods that learn to generate compounds with desired pharmacological properties. Nevertheless, RL-based methods typically condense the evaluation of sampled compounds into a single scalar value, making it difficult for the generative agent to learn the optimal policy. This work combines self-attention mechanisms and RL to generate promising molecules. The idea is to evaluate the relative significance of each atom and functional group in their interaction with the target, and to utilize this information for optimizing the Generator. Therefore, the framework for de novo drug design is composed of a Generator that samples new compounds combined with a Transformer-encoder and a biological affinity Predictor that evaluate the generated structures. Moreover, it takes the advantage of the knowledge encapsulated in the Transformer's attention weights to evaluate each token individually. We compared the performance of two output prediction strategies for the Transformer: standard and masked language model (MLM). The results show that the MLM Transformer is more effective in optimizing the Generator compared with the state-of-the-art works. Additionally, the evaluation models identified the most important regions of each molecule for the biological interaction with the target. As a case study, we generated synthesizable hit compounds that can be putative inhibitors of the enzyme ubiquitin-specific protein 7 (USP7).

Artificial intelligence for prediction of biological activities and generation of molecular hits using stereochemical information.

In this work, we develop a method for generating targeted hit compounds by applying deep reinforcement learning and attention mechanisms to predict binding affinity against a biological target while considering stereochemical information. The novelty of this work is a deep model Predictor that can establish the relationship between chemical structures and their corresponding [Formula: see text] values. We thoroughly study the effect of different molecular descriptors such as ECFP4, ECFP6, SMILES and RDKFingerprint. Also, we demonstrated the importance of attention mechanisms to capture long-range dependencies in molecular sequences. Due to the importance of stereochemical information for the binding mechanism, this information was employed both in the prediction and generation processes. To identify the most promising hits, we apply the self-adaptive multi-objective optimization strategy. Moreover, to ensure the existence of stereochemical information, we consider all the possible enumerated stereoisomers to provide the most appropriate 3D structures. We evaluated this approach against the Ubiquitin-Specific Protease 7 (USP7) by generating putative inhibitors for this target. The predictor with SMILES notations as descriptor plus bidirectional recurrent neural network using attention mechanism has the best performance. Additionally, our methodology identify the regions of the generated molecules that are important for the interaction with the receptor's active site. Also, the obtained results demonstrate that it is possible to discover synthesizable molecules with high biological affinity for the target, containing the indication of their optimal stereochemical conformation.

FSM-DDTR: End-to-end feedback strategy for multi-objective De Novo drug design using transformers.

The design of compounds that target specific biological functions with relevant selectivity is critical in the context of drug discovery, especially due to the polypharmacological nature of most existing drug molecules. In recent years, in silico-based methods combined with deep learning have shown promising results in the de novo drug design challenge, leading to potential leads for biologically interesting targets. However, several of these methods overlook the importance of certain properties, such as validity rate and target selectivity, or simplify the generative process by neglecting the multi-objective nature of the pharmacological space. In this study, we propose a multi-objective Transformer-based architecture to generate drug candidates with desired molecular properties and increased selectivity toward a specific biological target. The framework consists of a Transformer-Decoder Generator that generates novel and valid compounds in the SMILES format notation, a Transformer-Encoder Predictor that estimates the binding affinity toward the biological target, and a feedback loop combined with a multi-objective optimization strategy to rank the generated molecules and condition the generating distribution around the targeted properties. The results demonstrate that the proposed architecture can generate novel and synthesizable small compounds with desired pharmacological properties toward a biologically relevant target. The unbiased Transformer-based Generator achieved superior performance in the novelty rate (97.38%) and comparable performance in terms of internal diversity, uniqueness, and validity against state-of-the-art baselines. The optimization of the unbiased Transformer-based Generator resulted in the generation of molecules exhibiting high binding affinity toward the Adenosine A2A Receptor (AA2AR) and possessing desirable physicochemical properties, where 99.36% of the generated molecules follow Lipinski's rule of five. Furthermore, the implementation of a feedback strategy, in conjunction with a multi-objective algorithm, effectively shifted the distribution of the generated molecules toward optimal values of molecular weight, molecular lipophilicity, topological polar surface area, synthetic accessibility score, and quantitative estimate of drug-likeness, without the necessity of prior training sets comprising molecules endowed with pharmacological properties of interest. Overall, this research study validates the applicability of a Transformer-based architecture in the context of drug design, capable of exploring the vast chemical representation space to generate novel molecules with improved pharmacological properties and target selectivity. The data and source code used in this study are available at: https://github.com/larngroup/FSM-DDTR.