Recurrent Neural Network Model for Constructive Peptide Design

Design of metalloproteins and novel protein folds using variational autoencoders

The design of novel proteins has many applications but remains an attritional process with success in isolated cases. Meanwhile, deep learning technologies have exploded in popularity in recent years and are increasingly applicable to biology due to the rise in available data. We attempt to link protein design and deep learning by using variational autoencoders to generate protein sequences conditioned on desired properties. Potential copper and calcium binding sites are added to non-metal binding proteins without human intervention and compared to a hidden Markov model. In another use case, a grammar of protein structures is developed and used to produce sequences for a novel protein topology. One candidate structure is found to be stable by molecular dynamics simulation. The ability of our model to confine the vast search space of protein sequences and to scale easily has the potential to assist in a variety of protein design tasks.

Artificial Intelligence in Drug Design

Artificial Intelligence (AI) plays a pivotal role in drug discovery. In particular artificial neural networks such as deep neural networks or recurrent networks drive this area. Numerous applications in property or activity predictions like physicochemical and ADMET properties have recently appeared and underpin the strength of this technology in quantitative structure-property relationships (QSPR) or quantitative structure-activity relationships (QSAR). Artificial intelligence in de novo design drives the generation of meaningful new biologically active molecules towards desired properties. Several examples establish the strength of artificial intelligence in this field. Combination with synthesis planning and ease of synthesis is feasible and more and more automated drug discovery by computers is expected in the near future.

DeepDTA: deep drug-target binding affinity prediction

Motivation

The identification of novel drug-target (DT) interactions is a substantial part of the drug discovery process. Most of the computational methods that have been proposed to predict DT interactions have focused on binary classification, where the goal is to determine whether a DT pair interacts or not. However, protein-ligand interactions assume a continuum of binding strength values, also called binding affinity and predicting this value still remains a challenge. The increase in the affinity data available in DT knowledge-bases allows the use of advanced learning techniques such as deep learning architectures in the prediction of binding affinities. In this study, we propose a deep-learning based model that uses only sequence information of both targets and drugs to predict DT interaction binding affinities. The few studies that focus on DT binding affinity prediction use either 3D structures of protein-ligand complexes or 2D features of compounds. One novel approach used in this work is the modeling of protein sequences and compound 1D representations with convolutional neural networks (CNNs).

Results

The results show that the proposed deep learning based model that uses the 1D representations of targets and drugs is an effective approach for drug target binding affinity prediction. The model in which high-level representations of a drug and a target are constructed via CNNs achieved the best Concordance Index (CI) performance in one of our larger benchmark datasets, outperforming the KronRLS algorithm and SimBoost, a state-of-the-art method for DT binding affinity prediction.

Availability and implementation

https://github.com/hkmztrk/DeepDTA.

Supplementary information

Supplementary data are available at Bioinformatics online.

Perturbation-Theory and Machine Learning (PTML) Model for High-Throughput Screening of Parham Reactions: Experimental and Theoretical Studies

Machine learning (ML) algorithms are gaining importance in the processing of chemical information and modeling of chemical reactivity problems. In this work, we have developed a perturbation-theory and machine learning (PTML) model combining perturbation theory (PT) and ML algorithms for predicting the yield of a given reaction. For this purpose, we have selected Parham cyclization, which is a general and powerful tool for the synthesis of heterocyclic and carbocyclic compounds. This reaction has both structural (substitution pattern on the substrate, internal electrophile, ring size, etc.) and operational variables (organolithium reagent, solvent, temperature, time, etc.), so predicting the effect of changes on substrate design (internal elelctrophile, halide, etc.) or reaction conditions on the yield is an important task that could help to optimize the reaction design. The PTML model developed uses PT operators to account for perturbations under experimental conditions and/or structural variables of all the molecules involved in a query reaction, compared to a reaction of reference. Thus, a dataset of >100 reactions has been collected for different substrates and internal electrophiles, under different reaction conditions, with a wide range of yields (0-98%). The best PTML model found using General Linear Regression (GLR) has R = 0.88 in training and R = 0.83 in external validation series for 10 000 pairs of query and reference reactions. The PTML model has a final R = 0.95 for all reactions using multiple reactions of reference. We also report a comparative study of linear versus nonlinear PTML models based on artificial neural network (ANN) algorithms. PTML-ANN models (LNN, MLP, RBF) with R ≈ 0.1-0.8 do not outperform the first PMTL model. This result confirms the validity of the linearity of the model. Next, we carried out an experimental and theoretical study of nonreported Parham reactions to illustrate the practical use of the PTML model. A 500 000-point simulation and a Hammett analysis of the reactivity space of Parham reactions are also reported.