Transformer neural network for protein-specific de novo drug generation as a machine translation problem
Трансформерная нейронная сеть для специфичной к белку генерации лекарственных соединений de novo как задача машинного перевода
2021-01-11
SCID: 54.1/xyzuc735
Discuss with AI
SMILES representationTransformer neural networkde novo drug designprotein amino acid sequenceself-attention
Figures from the paper
Abstract (AI)
Drug discovery for a protein target is a very laborious, long and costly process. Machine learning approaches and, in particular, deep generative networks can substantially reduce development time and costs. However, the majority of methods imply prior knowledge of protein binders, their physicochemical characteristics or the three-dimensional structure of the protein. The method proposed in this work generates novel molecules with predicted ability to bind a target protein by relying on its amino acid sequence only. We consider target-specific de novo drug design as a translational problem between the amino acid "language" and simplified molecular input line entry system representation of the molecule. To tackle this problem, we apply Transformer neural network architecture, a state-of-the-art approach in sequence transduction tasks. Transformer is based on a self-attention technique, which allows the capture of long-range dependencies between items in sequence. The model generates realistic diverse compounds with structural novelty. The computed physicochemical properties and common metrics used in drug discovery fall within the plausible drug-like range of values.
Key Findings
1
A Transformer neural network with self-attention is applied to capture long-range dependencies in protein and molecular sequences.
2
Generated molecules exhibit plausible drug-like physicochemical properties and values for commonly used drug-discovery metrics.
3
Target-specific molecular design is formulated as translation between the protein amino acid sequence and the molecule’s SMILES representation.
4
The method performs target-specific de novo drug generation using only a protein’s amino acid sequence, without requiring known binders, physicochemical profiles, or three-dimensional structure.
5
The model generates realistic, diverse, and structurally novel compounds predicted to bind the target protein.
Research Object
protein-target-specific de novo drug generation, represented by amino acid sequences and generated molecular SMILES
Research Subject
the ability of a Transformer-based sequence translation model to generate structurally novel, diverse, and drug-like molecules predicted to bind the target protein using only its amino acid sequence
Publication Details
Publication Date
2021-01-11
Journal
Publisher
ISSN
Cited by
266
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai4
Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation2014
Proceedings of the 24th international conference on Machine learning2007
Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules2018
State-of-the-art augmented NLP transformer models for direct and single-step retrosynthesis2020