We Ask Better Questions Than We Can Measure

Scientists' abilities to open new frontiers in cell biological research continue to be limited by current technologies rather than by questions or hypotheses. There are clearly defined gaps in measurement capability: we cannot image most cellular processes at the spatial and temporal resolution needed, cannot measure multiple molecular species simultaneously in living cells, cannot track single molecules over biologically relevant timescales, and cannot probe cell function without perturbing it. Biological processes at all scales — from molecular interactions to ecosystem dynamics — are determined through encoding, exchange, and interpretation of information, but our capacity to acquire, manage, analyze, and represent biological information is insufficient for the complexity of the questions being asked.

Every field of biology is ultimately limited by what can be measured. Cancer biology needs tools to track single-cell heterogeneity in tumors in real time. Neuroscience needs tools to record from all neurons in a circuit simultaneously. Developmental biology needs tools to follow individual cells through an embryo's development without disrupting it. Synthetic biology needs tools to measure metabolic flux in engineered organisms. These are not niche requirements — they represent the frontier of multiple $1B+ research programs worldwide. The gap between what biologists need to measure and what they can measure has widened as questions have become more sophisticated while instrumentation has evolved incrementally.

Fluorescence microscopy (the workhorse of cell biology) is limited to 3–4 simultaneous channels, requires fluorescent labels that can perturb cell function, and causes phototoxicity during long-term imaging. Super-resolution microscopy can see below the diffraction limit but trades spatial resolution for temporal resolution, making it unsuitable for dynamic processes. Cryo-electron microscopy provides atomic-resolution structures but only of frozen, dead samples — it cannot capture dynamics. Mass spectrometry-based proteomics is destructive (requires cell lysis) and loses spatial information. Single-cell RNA sequencing captures transcriptomes of individual cells but requires destroying the cell and provides only a snapshot, not a time series. Tools developed for specific labs often lack the robust engineering, documentation, and support needed for broad adoption by the research community, creating a "last mile" problem in tool dissemination.

Novel instrumentation addressing clearly defined gaps in observing biological phenomena — particularly non-destructive, multiplexed, long-duration measurement at single-cell and subcellular resolution. Interdisciplinary approaches drawing on advances from chemistry (novel probes), computer science (computational imaging, AI-enhanced reconstruction), engineering (microfluidics, MEMS), mathematics (compressed sensing, inverse problems), and physics (quantum sensing, advanced optics) applied to cell biology. Infrastructure that bridges tool development and community adoption — moving beyond "works in my lab" to broadly deployable instruments and methods.

A student team could develop a computational imaging pipeline that extracts more information from standard fluorescence microscopy data — for example, using machine learning to predict additional cellular features from existing image channels (virtual staining) or to deconvolve temporal dynamics from snapshot data. Alternatively, a team could design and fabricate a microfluidic device for automated, long-term single-cell culture with integrated optical measurement, using low-cost fabrication methods (3D printing, laser cutting). Relevant skills: optics, image processing, machine learning, microfluidics, cell biology, engineering.