The Hormoz Lab
Read the past.
Learn the rules.
Control the future.
Every biological system has a history, and that history shapes what it can become. We invent technologies to make these histories observable, develop mathematical models to discover how history determines fate, and use those discoveries to control cells, populations, and developing tissues.
Based at
Histories unfold across time
How we work
History becomes useful when it reveals the rules of change.
Make histories observable
We engineer cells to record their experiences, watch cellular families across generations, and reconstruct the past from molecular traces.
Infer how history determines fate
We develop mathematics that reveals hidden states, transition rules, clonal fitness, and the local interactions that build tissues.
Control what happens next
We perturb biological systems to test the inferred rules, then use those rules to redirect cells, populations, and developing tissues.
Current questions
What histories can biology remember?
01Synthetic biology · Lineage recording
Can cells record their own past?
We engineer cells and mice that write lineage and signaling histories into their own genomes.
Explore project
02Live imaging · Stochastic processes
Can family trees reveal invisible cell states?
We follow mammalian cells for many generations and infer hidden states from inherited behavior.
Explore project
03Cancer genomics · Population genetics
Can present-day blood reveal when cancer began?
We reconstruct the early history of blood cancers from naturally occurring somatic mutations.
Explore project
04Development · Synthetic embryos
Can we recreate and control human development?
We build synthetic embryos and learn which sequences of signals produce tissues and blood stem cells.
Explore project
05Single-cell genomics · Bacteria
What is a cell state in a bacterium?
We measure and perturb individual bacteria to learn why genetically identical cells choose different states.
Explore project
06Experimental evolution · HIV
Can we predict evolution from viral genomes?
We follow complete viral genomes through experimental evolution and infer the landscapes that constrain their futures.
Explore project
Mentoring in the lab
Mentoring and lab culture
“I see mentoring as a gradual transfer of scientific independence.”
Our lab is for scientists who want to cross boundaries between biology, engineering, mathematics, and computation. We do not expect students to arrive as experts in every area. We look for curiosity, rigor, and a desire to learn.
How we work togetherSelected work
Recent ideas and technologies
arXiv2025
DiffeoMorph: Learning to Morph 3D Shapes Using Differentiable Agent-Based Simulations
A differentiable agent-based model learns local rules that cause cells to assemble into target three-dimensional shapes.
ICLR2025
Improving Graph Neural Networks by Learning Continuous Edge Directions
A graph neural network learns the direction of information flow directly from data.
IEEE Transactions on Signal Processing2025
An efficient solution to Hidden Markov Models on trees with coupled branches
New mathematics for inferring hidden dynamics on branching cellular lineages.
2026–2027 rotations
Build experiments and theory together.
We are recruiting Harvard graduate students who want to learn how biological history determines what happens next.
See rotation projects