Accutar Biotech – Harnessing AI and Computational Technology to Revolutionize Drug Discovery and Design

The primary goal of drug discovery research is to identify medicines that have a beneficial effect on the body – in other words, medications that can help prevent or treat a specific disease. Although there are many different types of drugs, many are small chemically synthesized molecules. These are capable of binding specifically a target molecule present in the disease. Traditional drug discovery methods are target-driven, which means that a known target is used to screen for small molecules that interact with it or affect its function in cells. These approaches work well for easily druggable targets with a well-defined structure and well-understood intracellular interactions. However, due to the complexity of cellular interactions and a lack of knowledge of intricate cellular pathways, these methods are minimal.

AI in drug discovery market can overcome these obstacles by detecting novel interactions and determining the functional significance of various cellular pathway components. AI uses complex algorithms and machine learning to extract meaningful information from large datasets. For example, AI uses RNA sequencing data to identify genes whose expression correlates with a given cellular condition. AI can also identify compounds that could bind to ‘undruggable targets' or proteins with undefined structures. A predictive set of compounds can be easily placed in a relatively short amount of time by iterative simulations of interactions of different compounds with small pieces of a protein. With that, it is not surprising that experts are now looking to AI systems' unparalleled data processing potential as a way to accelerate and lower the cost of discovering new drugs.

Accutar Biotech is one such firm that employs artificial intelligence to revolutionize drug discovery. With capabilities in side chain flexible mode ligand docking, virtual screening, and drug ADME property prediction, Accutar’s platform beats the industry standard in computation-aided drug design. The company’s hybrid based approach, which uses computational drug design followed by wet lab validation, greatly reduces the time and cost necessary for traditional drug discovery efforts. Accutar is committed to building strong partnerships, and collaborates across academia and the pharmaceutical industry to solve interdisciplinary problems.

Leveraging Cutting-Edge AI-Based Drug Discovery Tools

ChemiRise: Accutar has developed an end-to-end, retrosynthesis system, named ChemiRise that can propose complete retrosynthesis routes for organic compounds rapidly and reliably. The system was trained on a processed patent database of over 3 million organic reactions. Experimental reactions were atom-mapped, clustered, and extracted into reaction templates. Accutar then trained a graph convolutional neural network-based one-step reaction proposer using template embeddings and developed a guiding algorithm on the directed acyclic graph (DAG) of chemical compounds to find the best candidate to explore. The atom-mapping algorithm and the one-step reaction proposer were benchmarked against previous studies and showed better results. The final product was demonstrated by retrosynthesis routes reviewed and rated by human experts, showing satisfying functionality and a potential productivity boost in real-life use cases.

Chemi-net: It is a graph convolutional network for accurate drug property prediction. Absorption, distribution, metabolism, and excretion (ADME) studies are critical for drug discovery. Conventionally, these tasks, together with other chemical property predictions, rely on domain-specific feature descriptors, or fingerprints. Following the recent success of neural networks, the firm developed Chemi-Net, a completely data-driven, domain knowledge-free, deep learning method for ADME property prediction. To compare the relative performance of Chemi-Net with Cubist, one of the popular machine learning programs used by Amgen, a large-scale ADME property prediction study was performed on-site at Amgen. The results showed that their deep neural network method improved current methods by a large margin. Accutar foresees that the significantly increased accuracy of ADME prediction seen with Chemi-Net over Cubist will greatly accelerate drug discovery.

Orbital: Orbital is a deep neural network-based docking platform. Simply provide the ligand in simplified molecular-input line-entry system (SMILES) format and the target protein, and docking becomes as easy as “one click” as demonstrated in the video below. The typical ligand preparation and target grid definition, hydration, and other parameter-setting steps are no longer needed, because the traditional force field-based energy equations method was abandoned in their approach. Instead, a dynamic deep neural network framework performs the docking task with unprecedented accuracy and speed. The average running time for a drug docking case using “dock all” mode scanning for all potential binding pockets of a target protein is under 1 minute on a standard laptop.

AC0682: It is an investigational orally bioavailable, chimeric degrader of estrogen receptor (ER) α for the potential treatment of ER-positive / human epidermal growth factor receptor 2 (HER2)-negative breast cancers. In preclinical studies, AC0682 has demonstrated potent and selective protein degradation of ERα wildtype and mutants with favorable pharmacological properties and brain penetration, as well as promising anti-tumor activities in ER-positive animal tumor models. AC0682 offers a potential new breast cancer treatment based on a differentiated mechanism of action from fulvestrant and novel SERDs. The purpose of the Phase 1 multi-center, open-label study is to assess the safety, tolerability, pharmacokinetics, and preliminary anti-tumor activity of AC0682 treatment in patients with ER-positive / HER2-negative locally advanced or metastatic breast cancer.

Meet the Leader

Jie Fan, PhD is the Founder and Chief Executive Officer of Accutar Biotech. Dr. Jie Fan was trained at the University of California, Berkeley in biostatistics and obtained his Doctorate degree from the Weill Cornell Graduate School of Medical Sciences and the Memorial Sloan Kettering Cancer Center in structural biology and immunology. He received further training with the Nobel Prize-winning biologist, Dr. Günter Blobel, at the Rockefeller University in New York. With the dream of using a hybrid approach, combining computation design and experimental validation, to accelerate drug discovery and aiming to reform the current “hit-to-lead” drug discovery process, Jie founded Accutar Biotech with the support of Dr. Günter Blobel.

“We aim to create and work from a data-driven foundation. We use a hybrid approach – computational drug design followed by wet lab validation – to greatly reduce the time and cost necessary for traditional drug discovery efforts.”