Abstract
This paper is a response to the increasing difficulty biologists find in agreeing upon a definition of the gene, and indeed, the increasing disarray in which that concept finds itself. After briefly reviewing these problems, we propose an alternative to both the concept and the word gene—an alternative that, like the gene, is intended to capture the essence of inheritance, but which is both richer and more expressive. It is also clearer in its separation of what the organism statically is (what it tangibly inherits) and what it dynamically does (its functionality and behavior). Our proposal of a genetic functor, or genitor, is a sweeping extension of the classical genotype/phenotype paradigm, yet it appears to be faithful to the findings of contemporary biology, encompassing many of the recently emerging—and surprisingly complex—links between structure and functionality.
The Crisis of the Gene Concept
For nearly a century, the "gene" served as the fundamental unit of inheritance—a discrete "bead on a string" that coded for a single protein. However, as genomic research has advanced, this classical definition has essentially collapsed. Evelyn Fox Keller and David Harel argue that the term has become a "crude four-letter word" that hinders scientific precision due to two primary factors:
- Genomic Complexity: We now understand that 98-99% of the human genome is non-coding, yet much of it remains functional. Sequences overlap, are spliced in thousands of different ways, and act as regulatory switches rather than simple protein templates.
- Conceptual Confusion: The word "gene" attempts to describe both what a sequence is (structure) and what it does (function) simultaneously. When a single stretch of DNA performs multiple functions across different contexts, the "gene" label loses its utility.
The Genitor Framework
To resolve this "disarray," the authors propose a formal logical framework rooted in computer science formalisms. This framework decouples the physical sequence from its functional outcome through three core components:
- Dene (D): A logical predicate or property of a DNA sequence. It is static and describes the physical structure (e.g., "The sequence contains motif X followed by motif Y").
- Bene (B): A logical predicate describing a biological behavior or "trace" over time. It is dynamic and describes functionality (e.g., "The organism produces protein Z within two minutes of an environmental trigger").
- Genitor (G): A triple $(O, D, B)$ representing the relationship: "In organism $O$, if the DNA satisfies property $D$, then the organism will exhibit behavior $B$."
By treating biology as a reactive system, the Genitor framework allows for the application of formal logic and verification techniques to model life more accurately than a simple list of genes.
Addressing Multiplicity and Context
One of the primary strengths of the Genitor model is its ability to handle the "Library of Babel" nature of the genome. Borrowing from Jorge Luis Borges, the authors view DNA not as a "Book of Life" with a fixed narrative, but as a library of all possible structures where the cell "reads" and "articulates" specific meanings based on context.
In this framework, multiplicity is a feature rather than a bug. A single nucleotide can be part of a thousand different Genitors, resolving the difficulty of defining where one gene ends and another begins. This accounts for:
- Alternative Splicing: Different Genitors can share the same DNA (Dene) but result in different behaviors (Benes).
- Non-coding RNA and Epigenetics: Regulatory sequences and heritable chromatin marks are easily incorporated as properties of the sequence or the organismal context.
Executability and Future Directions
The authors argue that biological models should be "playable." If a Genitor is formally defined, researchers should be able to simulate changes in the DNA sequence and observe the resulting shifts in behavior. This "executability" transforms biological modeling from a descriptive exercise into a predictive, computational one.
However, significant challenges remain. The specific formal languages (such as temporal or interval logic) best suited for defining Denes and Benes are still being refined. Furthermore, modeling an entire organism as a collection of Genitors represents a "Grand Challenge" for computing, requiring massive leaps in how we represent biological reactivity and the inheritance of the cellular machinery required to interpret these genetic instructions.
Conclusion
The Genitor framework offers a sweeping extension of the classical genotype/phenotype paradigm. By separating the static "Dene" from the dynamic "Bene," Keller and Harel provide a language capable of capturing the complex, modular, and context-dependent nature of modern genomics, moving biology beyond the limitations of the "gene."
Summary updated