SandboxAQ

SandboxAQ Essential Concepts: Quantum Gates A quantum gate is a fundamental building block of quantum computing, manipulating the state of a qubit to perform a computation. Unlike classical logic gates, quantum gates are reversible and can create superposition and entanglement, the essential ingredients that give quantum computers their unique capabilities. Our latest Essential Concepts visual breaks down the most common quantum gates, how they work together in a quantum circuit, and how these operations are physically performed on different types of quantum hardware. Read through it below and share your thoughts.

A computational method is only as useful as its performance across the problems researchers actually face. This work comprehensively benchmarks AQFEP across diverse chemical systems, including TYK2, EGFR, PTP1B, CRBN, molecular glues, and GLP1R. The study also examines how Bayesian optimization can improve system-specific simulation settings and how AQFEP can support virtual screening at approximately 1-2 hours per ligand on a single T4 GPU. In one virtual-screening study, a library of 5 million commercially available compounds was narrowed to 1,457 candidates, producing a 2.7% hit rate and three advancing lead series. The takeaway: rigorous benchmarking matters because generalizability matters. The goal is not a method that works once, but a workflow that can help teams make better decisions across chemical space. Interested about learning more about our AI Scientific Discovery Summit on August 27th? https://lnkd.in/eVwceTAN

In the latest fAQ podcast, host Tai-Danae Bradley sits down with Janet Faakye, Ph.D., PMP, neuroscientist and project manager at SandboxAQ, to talk about her journey from Ghana, where she studied psychology, to a life-changing course in neurobiology that altered her career path and brought her to America. Together they begin to answer the question, what happens when a passion for science meets a need to make science understood? Janet shares what studying brain-aging taught her about communicating complex science, how a SandboxAQ Ph.D residency opened the door to a career in deep tech, why supporting and validating the next generation of scientists matters, and more! Janet’s story is a great reminder that there’s no single path into science or deep tech. We think it's a must-listen for your week: https://lnkd.in/d9bFuCQh

SandboxAQ Essential Concepts – On Measurement: Why Looking Changes Everything What if the simple act of measuring something could actually change it? In classical physics, measurement barely disturbs a system. In quantum physics, measurement can fundamentally change the state of a system. A qubit in superposition can exist in a combination of states, but once it’s measured, that superposition collapses into a single state. We created this visual to break down the difference between classical and quantum measurement, and why timing, isolation, and precision matter so much in quantum computing. If you’ve ever wondered why measurement is such a critical part of the quantum world, this is a great reference.

Ternary complexes are central to targeted protein degradation, and they are notoriously hard to predict. The relevant biology depends on more than a ligand binding a single protein; it depends on how the ligand, target protein, and E3 ligase come together in three-dimensional space. This work combines cofolding with AQFEP calculations to improve binding-affinity prediction for E3 ligase modulators. Algorithmic truncation helps address structural interference in ternary-complex predictions, while Bayesian optimization identifies more effective, system-specific simulation settings. The workflow produces predictions in approximately 1–2 hours per ligand on a single T4 GPU, creating a practical path toward scaling computational support for protein-degradation campaigns. The bigger opportunity is to make complex molecular systems more tractable for hit discovery and lead optimization. Interested in joining us at our AI Scientific Discovery Summit on August 27th in Boston? Request to attend here: https://lnkd.in/eVwceTAN

Binding affinity matters, but for GPCR drug discovery it only tells part of the story. A compound can bind to a receptor and still produce very different functional outcomes depending on which receptor state it stabilizes. In this work, we combine machine-learning affinity prediction with physics-based AQFEP calculations across modeled active and inactive GPCR states. The result is a workflow that supports both efficient ligand prioritization and more mechanism-aware prediction of agonist versus antagonist behavior. In retrospective benchmarking, the ML model prioritized approximately 22% of the library with 79% accuracy and 83% specificity, while AQFEP provided complementary sensitivity for binder discovery. The broader goal: move from asking “does it bind?” to asking “what does it do?” We’re looking forward to sharing this work at our upcoming event in Boston on August 27th. Request to attend here: https://lnkd.in/eVwceTAN

SandboxAQ Essential Concepts – Quantum Computers What if there wasn't just one way to build a quantum computer? Unlike classical computers, which rely on a standardized architecture, quantum computers can be built using several fundamentally different types of qubits. Each approach has its own strengths and engineering challenges. This visual highlights seven of the leading quantum computing platforms: superconducting circuits and trapped ions to photonic systems, neutral atoms, topological qubits, semiconductor quantum dots, and solid-state defects. While they operate in different ways, they all share the same goal: creating powerful quantum systems capable of solving problems beyond the reach of today's computers. Explore the different approaches driving the future of quantum computing. 👇

Finding a promising target is not just a structure-prediction problem. It is a context problem. This work integrates AI-based protein-structure prediction, proteochemometric modeling, knowledge graphs, and physics-based binding-affinity calculations into an in-silico target-identification workflow. The approach combines structure curation, binding-site and pose assessment, molecular-dynamics refinement, and AQFEP calculations with broader biological context from knowledge graphs. In inverse-screening experiments, true targets ranked as high as number one, with additional true targets appearing at numbers five and fifteen for one ligand, and at number one for another. By combining molecular detail with biological context, the workflow can support target prioritization across applications such as drug repurposing, toxicity, and selectivity prediction. Interested in joining us at our AI Scientific Discovery Summit on August 27th in Cambridge, MA? Learn more and request to attend here: https://lnkd.in/eVwceTAN

SandboxAQ Essential Concepts – The Founders of Quantum Mechanics Every field has its pioneers. Quantum mechanics had three who fundamentally changed how we understand reality. This infographic explores how: 🟡 Planck showed that energy comes in discrete packets, or quanta. 🟡 Einstein demonstrated that light behaves as particles called photons, explaining the photoelectric effect. 🟡 Bohr proposed that electrons occupy specific quantized energy levels within atoms. Together, these discoveries transformed our understanding of matter, light, and energy, and became the foundation for today's quantum technologies, including advanced computing, secure communications, new materials, and precision sensing. Take a look at the infographic below, and let us know: Which quantum pioneer has had the biggest impact on you or your research?