What this covers

Rosemary Braun, a biophysicist at Northwestern University, joins Sean Carroll to discuss how organisms function as integrated wholes rather than collections of independent parts. The episode centers on what large-scale biological datasets reveal about coordination and control in living systems. Braun argues that while individual molecules and cells follow local physical rules without awareness of the broader organism, the system as a whole exhibits robust, purposeful behavior. The conversation ranges across how cells operate on low-dimensional manifolds, why reductionist approaches miss crucial dynamics, and how circadian biology—timekeeping at the cellular level—orchestrates physiology without centralized command.

The episode takes up several concrete domains where systems-level thinking reshapes understanding. Nearly every cell contains a core clock, a molecular oscillator with roughly a 24-hour period built from negative feedback with time delay, and this circadian mechanism regulates a large fraction of gene expression across the body. Braun discusses evidence that these clocks coordinate metabolism and lifespan without direct cell-to-cell communication, and explores how timing of food intake and exercise can shift circadian phase—with implications for treating jet lag and administering drugs at optimal times. The conversation also covers why microRNAs, which bind multiple targets non-specifically, may constitute a mechanism for systems-level control rather than precision targeting of individual genes. A persistent tension surfaces throughout: biologists understand that organism-level outcomes depend on network properties, yet experimental tools only act on individual components, which keeps reductionist intuition alive even when unhelpful.

Sharpest takeaway

Braun argues that biological organisms are best understood not by reducing them to individual genes but as emergent systems whose reliable macroscopic behavior arises from low-dimensional, coarse-grained, systems-level patterns of interaction among thousands of components.

  • The vast omics data lives on a low-dimensional manifold, so only a few collective variables actually govern cell behavior
  • Regulatory molecules like microRNAs act non-specifically to exert systems-level rather than one-to-one control
  • Circadian clocks exist in nearly every cell and coordinate processes across the organism without direct communication

The argument · threads4 threads · 33 claims
0.85

Physics succeeds by tracking few variables, but biology lacks obvious measurable variables like physics does.

1 pointcentrality 5/5
  • To figure out how the earth orbits the sun, we do not need to know what the earth's 10^50 atoms are doing—we only need a small number of variables like the center of mass and velocity—and in physics it is fairly obvious which variables we need and they coincide with what we can easily measure, unlike biology.

    “to figure out how the earth orbits the sun, we don't need to know what those atoms are doing. We just need to know a small number of variables, the center of mass and the velocity.”

0.71

Biological parts lack clear single purposes because selection optimized whole systems, not individual components.

1 pointcentrality 4/5
  • Our intuition that each part has a clear single purpose comes from designed systems, but because biological organisms arose through natural selection where the whole system was successful, the boundaries and functions of parts like genes and proteins are often not clearly delineated.

    “because all the stuff that we see in biological organisms have just sort of come about because the whole system was successful the boundaries are not always so clear in biology.”

0.71

Biological resilience and wide developmental basins contrast with mechanical brittleness and fine-tuning requirements.

1 pointcentrality 4/5
  • Mechanical systems tend to be brittle—break a single wire and the computer stops working—whereas biology has to be resilient, which goes hand in hand with development having a wide basin of attraction rather than being fine-tuned.

    “Mechanical things tend to be brittle, right? You break a single wire and then your computer doesn't work, but biology has to be a little resilient.”

0.44

More data is always better in biology.

1 pointcentrality 2/5
  • More data is always better in biology.

    “more data isn't, I don't wanna say it's not always better, more data is always better, but it's not useful by itself”