Gene Switches and Decision Trees Determine our Development

Philip Iannaccone
3 min readSep 11, 2018


Genes can be turned on and off just like switches. The switches are complex little machines with many pieces that we are only now just beginning to understand. Imagine confronting a light switch with no knowledge of how it works only that it makes lights go on and off. If you were a curious person you might try to figure out how it worked.

But what would you do?

More than likely you would start by taking it apart and see the internal parts and how they fit together. Then you might start removing parts to determine their effect on the predicted outcome (lights going on or off). You might try to put it back together and see if still works. Encouraged with this knowledge you might work out ways to fix broken switches. Ultimately you might think “I can do better” and try to design your own switch and build it from scratch based on what you have learned. Developmental geneticists do the same thing.

Many years of research has lead to the realization that much of early developmental decisions from fertilization to birth and after are the consequence of a cell’s decision to activate one of two key transcription factors (the little molecular switches) at critical points in time during development.

At different critical junctures during development different pairs of transcription factor are vying with each other. It is like the toy below. In this model there is a bunch of marbles that are all alike. Then there are little gates with a switch. If the switch is one position the marble goes one way, if it is in the other position the marble goes in the other direction. This decision tree extends four levels with each level doubling the number of pathways that can be selected depending on where the switch is set.

Development proceeds as a series of decisions many of which behave like binary switches. If the cells that make up developing tissues are represented by the colored marbles at the top, the switches on the left determine which path the marble will take. The switches are set in response to embryonic signals controlling gene expression. In total each decision defines the proper bin at the bottom that represents the differentiated state necessary to organize the tissue. If a switch is incorrectly set, the state of the cell is wrong and it cannot respond properly to the next set of signals. The result is that the cell winds up in the wrong bin at the bottom. Since the switches following the mistake are responding to the next developmental signal in an unknown manner the bins the cells wind up in and thus the tissue aberration is difficult if not impossible to predict.

At the end of this simple binary decision tree 16 different fates arise as the different slots at the bottom.

How much more powerful would the tree become with the addition of more switches?

More and more branching is how we wind up as so many different types of cells.

But if a mistake is made and the marble goes down an arrant pathway it can’t get back. In terms of the development of the fetus this would result in what physicist Murray Gell-Mann called “frozen accidents” in describing why the universe wound up as it is. After the mistake the resulting misdirection means the wrong cells (marbles) are in the wrong place at the next juncture. So the next decision can’t be correct and can’t be reversed resulting in incorrect tissue and signaling components in the wrong place and at the wrong time. The results of these incorrect decisions are often unexpected, unpredictable, and confusing.

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.

Originally published at



Philip Iannaccone

Phil Iannaccone is a Professor of Pediatrics and Pathology at Northwestern University Feinberg School of Medicine.