My mission is to reduce human suffering. We accomplish this using biomedical data science.

I'm an Assistant Professor in the Department of Biomedical Informatics and a member of the Center for Health AI at the University of Colorado Anschutz, where I am the Principal Investigator of the Way Science Lab.

Cell morphology encodes disease state. By comparing high-content cell morphology profiles of healthy (wild-type) cells against diseased cells — for example, identifying the phenotypic signature that distinguishes NF1-null Schwann cells from wild-type (Tomkinson et al. 2025) — we define phenotypic targets. This approach is complementary to gene-target-based drug discovery, and may better capture where disease actually manifests at the cellular level. We then screen large compound libraries against these phenotypic targets to find drugs that restore healthy morphology.

My lab develops computational methods and open-source software to analyze high-content microscopy images of cells, partnering with collaborators worldwide to develop innovative assays, extract biological meaning, and advance drug discovery. Our disease focus spans:

• 🧬 NF1 personalized medicine — identifying morphology signatures of Neurofibromatosis Type 1 and running high-throughput drug screens to find compounds that restore a healthy phenotype for Neurofibromatosis Type 1 (NF1).
• ❤️ Cardiac fibrosis — high-throughput drug screening using morphology .profiles of human cardiac fibroblasts to prioritize therapeutic candidates
• 🎗️ Pediatric cancer — high-throughput phenotypic drug screening across multiple high-risk cancer types with few effective treatment options
• 🔬 Image-based profiling — methods for processing rich, reproducible biological signals from microscopy images at scale.
• 🤖 Machine learning for biology — models that connect cell morphology to genetic, molecular, and disease states to define phenotypic targets.
• 🛠️ Open-source software — reproducible tools that make high-content imaging data accessible to the broader research community.

Neurofibromatosis Type 1 (NF1) is a genetic disease impacting 1 in 2,500 people, with significant unmet therapeutic need. We use patient-derived organoid models combined with high-content microscopy and machine learning to identify morphological disease signatures and perform high-throughput drug screening — with the goal of enabling personalized medicine and drug discovery for NF1 patients.

The example 3D microscopy images below shows raw fluorescence channels alongside our segmentation masks for a single NF1 organoid:

DNA (405nm)	ER (488nm)	AGP (555nm)	Mitochondria (640nm)
Nuclei mask	Cell mask	Cytoplasm mask	Organoid mask

→ See the NF1 organoid profiling pipeline for full analysis code and data.

We extract 3D features from these organoids with:

We developed ZEDProfiler as a CPU-first toolkit for 3D image feature extraction built specifically for high-content NF organoid profiling. It handles multi-channel volumetric data (DNA, ER, AGP, mitochondria), anisotropic voxel spacing, and multiple segmentation compartments (nuclei, cytoplasm, whole cell, organoid) — computing the morphological measurements that feed directly into our drug screening pipeline.

I lead and maintain the Cytomining open-source ecosystem and am a member of the CytoData scientific community.

• 💬 Join us on Discord
• 🏫 Learn more at cytodata.org
• 🌈 We follow a Code of Conduct

Project		Description
	pycytominer	Python package for image-based profiling bioinformatics
	CytoTable	Harmonize high-content image analysis tool outputs
	coSMicQC	Single-cell morphology quality control
	buscar	Perturbation hit calling for high-content screening

→ See the full Cytomining ecosystem for more tools and projects.

Co-founded with Dave Bunten, Software Gardening is an ecosystem of applied guidance and tools for sustainable software development and maintenance in scientific computing. Software Gardening is supported by the Better Scientific Software program and the Sustainable Horizons Institute.

→ Explore the Software Gardening Almanack → Read the BSSw blog post

I have authored 50+ peer-reviewed publications, which include >10,000 citations, and an h-index of 25+, including work in Nature, Cell, Science, Nature Methods, Nature Communications, Circulation, PNAS, eLife, and Cell Systems. My research spans image-based profiling, machine learning for biology, cancer genomics, and open-source software.

→ Full publication list on Google Scholar

Ad-hoc reviewer for Nature, Science, PNAS, Cell, Nature Communications, Nature Methods, Nature Reviews Genetics, Cell Systems, Genome Medicine, Journal of the National Cancer Institute, Patterns, Bioinformatics, PLoS Computational Biology, and others.

This work is supported by:

Gilbert Family Foundation	Alex's Lemonade Stand Foundation	American Heart Association	National Institutes of Health