This session will be a unvarnished view on the current state of quantum computing in the pharma space. We’ll discuss examples of things that have potential, and then we’ll spend some time on hype-killing and dispelling illusions. We’ll talk more importantly about the path forward and the essential nature of collaboration to help keep the future aligned to the potential value and discuss the role that pharma companies need to play in that collaboration.

In this talk, I will discuss my research on using quantum computing and quantum neural networks for precision medicine. We are exploring two use cases: i) personalized treatment of knee osteoarthritis, and ii) neurodegenerative disease early diagnosis from brain MRI. I will briefly present our validation study “A quantum-enhanced precision medicine application to support data-driven clinical decisions for the personalized treatment of advanced knee osteoarthritis: development and preliminary validation of precisionKNEE_QNN” (recently included in the prestigious reviews from Oxford Academic) and discuss the potential impact of this work on the future of precision medicine. I will also discuss our ongoing R&D project on using quantum computing to predict effective brain age from MRI data for precision medicine in neurodegenerative diseases.

Quantum computing, a novel technology, introduces innovative algorithmic approaches that facilitate the execution of specific computations much more rapidly than classical computing. This presentation delves into the potential uses of quantum computing in the realms of healthcare and pharmaceuticals, encompassing drug discovery, personalized medicine, and medical imaging. Furthermore, we will outline the progression of the quantum ecosystem and address various challenges and constraints associated with quantum computing.

Polaris Quantum Biotech (POLARISqb), is developing QuADD (Quantum-Aided Drug Design), a subscription-based SaaS product that finds your lead-like hits from a library of 10^30 molecules (and growing) in 1-2 days and answers the question “When looking for new drugs, where do you start?" In this presentation, we will discuss the requirements and quantum technology behind QuADD, as well as several case studies and how you can test it.

Chugai has developed original mid-size molecular drug discovery technologies, which can generate orally bioavailable cyclic peptides as a new therapeutic modality. In this context, we are exploring invaluable applications of quantum computing technology to our mid-size molecular research. Based on benchmark survey with attending some quantum ecosystems, we are planning and conducting POC studies for examining its potential for enhancing our mid-size molecular research. As a case study, we will introduce a POC study of cyclic-peptide docking simulation using a quantum computing-inspired optimization solution.

I will present an accessible overview of two types of quantum algorithms which have been predicted to have scaling advantages over their classical counterparts. Of particular interest are algorithms that solve combinatorial optimization or compute chemical properties of molecules, and this presentation will discuss specific life science domain applications in the literature that fall into these two camps. Accessible to non-experts in quantum computing, the talk aims to provide a clear introduction and discussion of specific methods in quantum computing that are active areas of research for applications to computational biology.

Our research explores the use of quantum machine learning in the biomedical sciences, merging insights from two recent studies. Utilizing Ising-type machine learning algorithms, we effectively analyzed multi-omics cancer data and classified severe COVID-19 patients. These methods surpass traditional models in deciphering complex disease patterns on smaller training data sets, underscoring the potential of unconventional computing approaches to address the surge of genomic data and enhance precision medicine.

Cancer genomics represent a crucial element of precision oncology, where broad-panel genomic profiling of tumor tissue or blood is used to identify genomic biomarkers of resistance to various anti-cancer therapies. The laboratory reports resulting from these tests are routinely 20-50 pages long and impractical for integration into the workflow of a practicing oncologist. Therefore, we have developed an Quantum AI workflow to produce more informed and accurate outputs.

Even at smaller scales, the numerical parameters governing cellular regulatory systems span a finite, yet highly multi-dimensional space, where distinct dynamical behaviors give rise to a diverse array of phenotypes. The outcomes of an analysis using quantum computing centered on a complex three-component motif within the cellular regulatory system will be presented. This motif governs a wide spectrum of cellular functions and phenotypes, exhibiting variations under different conditions and treatments. This approach not only enabled the discovery of rare functions and concealed parameter subspaces, but also showcased the advantages of quantum computing, including its high performance and rapid computational capabilities.

This talk presents a novel approach to drug discovery, leveraging AI/ML and quantum computing to revolutionize the identification of molecular targets for cancer therapies. By automating and scaling the molecular design workflow, we aim to significantly shorten the drug development cycle, reduce costs, and enhance precision, paving the way for more effective cancer treatments.

Phenotypic screening is mostly done by high content screening, where millions of images need to be stitched together and required to process by machine learning algorithms, hence required high processing power like GPUs. Quantum computing is seen as potential accelerant of AI and has potential to enhance the ability to process and analyze large data sets while also presenting novel risks. Drug discovery productivity has been on the decline in recent years as “easy” drug targets diminish and pressure to identify best-in-class drugs continues. The traditional drug discovery paradigm requires knowledge of the specific target of interest for target-based high throughput screens. While this approach is well established there is an increased emphasis on screening for changes in disease-relevant phenotypes that may be modulated through unanticipated protein or pathway targets. As such, phenotypic screening has become a preferred method in drug discovery to find best-in-class drugs for unmet medical needs. In the Lead Discovery and Optimization (LDO) department of Bristol Myers Squibb, we have developed a phenotypic screening infrastructure centered around disease-relevant cellular models with appropriate hypotheses to guide high throughput phenotypic screening strategy. Many phenotypic screens require high-content imaging to identify cellular phenotypes indicative of compounds with potential therapeutic efficacy. To enable expedited analysis and identification of novel phenotypes and mechanism of action of active compounds, the LDO data science and research IT teams have established a pipeline to implement an unsupervised machine learning approach using convolutional neural networks.​ Similarly, an informatics-based approach has been developed to mine historical screens from internal and external sources to suggest prognostic targets. Ultimately, these approaches will be combined to enable unbiased and diverse hit selection in phenotypic screens to ensure the highest odds of identifying progressible chemical matter. The infrastructure assembled as part of this joint data science/IT initiative will enable successful capture of datasets, scaling of models, and model deployment for decision-making.