Dove Medical Press is pleased to invite you to submit your research to an upcoming Article Collection on "Is CADD ready to take a leap in the light of AI? Where are we?" in Drug Design, Development and Therapy.
Artificial intelligence (AI) in a research domain represents the theory and the computer technology development that helps to model the physical reality of the phenomenon under study. AI is capable of solving complex problems with higher accuracy and efficiency, such as robust language models like GPT, 3D protein model prediction tools like AlphaFold, and defeating champions in a board game like AlphaGo etc. The success of AI is well appreciated in our day-to-day life as well as in corporate businesses. However, its impact on computer aided drug design (CADD) is still awaited.
The field of AI in CADD is in its initial stage, where the development of a smart machine is difficult with the amount of data available. Numerous machine learning (ML) models have been built by implementing neural networks (NN), deep neural network (DNN), convolutional NN etc. to predict drug properties, active inhibitors and pharmacophore, and protein-inhibitor affinity - but the accuracy and specificity remains limited due to insufficient data. Recently, AlphaFold2 has been very successful in CASP14 for the prediction of single domain protein structure with experimental resolution (~2Å), which will be of great assistance to structure-based drug design. However, improvement in feature design and novel machine learning approaches will be required for the advancement of ligand-based design and CADD in general.
There has been significant effort in deploying AI for drug prediction and its target recognition of diseases. AI-based methods such as machine learning and deep learning have revolutionized the field of drug discovery with new algorithms such as back propagation for update of weights, context capture by the use of attention layer, and transformer’s encoder-decoder architecture for input reconstruction. Combined bioinformatics and machine learning have shown great potential in drug target prediction and computer drug design by enabling the analysis of the available amount of biological data and predicting the potential interactions between drugs and their targets, such as cancer, Alzheimer's disease, and other neurological disorders. The endeavour to improve our understanding of the drugs, their targets, and the disease phenotype, is still on and we have yet to build the right tools to solve one of the most complicated question of the biological science.
The aim of this article collection is to publish articles related to AI assisted structure-based and ligand-based drug design under the subtopics mentioned below.
1. Current state and future development of novel small molecule design with desired physico-chemical properties/activities by implementing state of the art technique, AI assisted de novo design, synthesis, pharmacophore modelling, and toxicity prediction.
2. Current and future developments in novel drug target prediction and small molecule drug development through combined machine learning, bioinformatics and CADD technologies.
3. Current state and future development of novel feature/descriptor engineering and deep learning model building for protein inhibitor binding energy prediction.
Please note that any submissions based on in-silico research require to be validated using appropriate in vitro or in vivo methods as per the journal Aims and Scope requirements of Drug Design, Development and Therapy.
Please submit your manuscript on our website, quoting the promo code XRMJW to indicate that your submission is for consideration in this Article Collection.
Disclosure Statement: Guest Advisors Dr. Huang, Dr. Haque, Dr. Wang and Dr. Naidu declare no conflict of interest regarding this work.
