Communicating via messages in multicellular systems

Both life and computing have evolved two types of complex transferable information. The meaning of one type ("code") is whatever it causes the receiver to do when it is "executed." The meaning of the other type (data) is up to the receiver to decide.

The second of the four principles of multicellular systems is that the receiver, not the sender, of a message must determine the meaning of the message. That principle is known in computing as polymorphic messaging -- different receiving cells (or computers) may interpret the message differently according to their specialized function. Messages from external sources must not be "executed" blindly by standard machinery in either a cell or a computer.  Thus messages must not contain executable code or DNA/RNA that directly causes the receiver to do what it's told to do.

The rejection of transferable code is necessary to protect cells in a multicellular organism from being hijacked by a virus.  Cells have many layers of defense against infection by external DNA or RNA, and computers have many (though inadequate at present) layers of defense against malware. Defenses aside, receiver-determined meaning is the only feasible messaging architecture in any system where the individual units are specialized. A non-omniscient sender cannot know how to direct the behavior of all specialized receivers. Instead, the specialized individuals know how to manage their own specialized functions and need only messages from the larger organism that provide information about present needs.

Biological and digital messages are transmitted by linear sequences of interchangeable elements -- "alphabets" if you will. Cells use messenger molecules constructed of chains of simple chemical subunits called nucleotides, labeled in shorthand as A, C, T, G and U. Computers use messages composed of sequences of bytes.

Life has evolved two generic sorts of long-chain molecules.  One, molecules of DNA or RNA, are long chains of nucleotides that primarily serve to carry the genetic 'program' of the cell. The others are proteins which are chains of amino acids that fold tightly upon themselves to form complex shapes that determine their functional/structural properties. Protein chains tend to be from a few dozen to a few thousand amino acids in length. Once folded, they become the "parts" that make up most of the machinery of the cell.  In contrast, DNA when not active in its "code" role, is folded tightly so that it cannot accidentally be "executed".  Only when playing its genetic coding role is it unfolded to expose its coding sequence to the protein machinery that interprets its execution. Chains of RNA can play both sorts of roles: they may fold into functional shapes, much like a protein, to act as "parts" in larger complexes, or their genetic sequence may be interpreted programmatically (the details of the two roles of RNA are beyond the scope of the current topic). In general, however, transfer of messenger proteins cause the cell to select behavior from its existing repertoire whereas transfer of genetic material changes the repertoire itself. Metazoan cells have predetermined functions and can seldom if ever tolerate having their functional repertoire changed.

Digital messages in computing are strings of bytes that range from idiosyncratic binary codes to highly structured XML messages such as REST, SOAP or other Web Services messages. Some strings are executable and some are not, depending on the computer's CPU (or interpreters for scripting languages such as Javascript or ActiveX). Thus, both life and computing have evolved two forms of complex information media: one executable and the other not.

The distinction between the two kinds of message is central to communication strategies in biology and communication strategies in computing. The parallels between the two realms can help us to understand multicellular computing. Whereas single-cells and single computers can afford to exchange executable code, and often benefit by doing so, code exchange in multicellular systems is exceedingly dangerous -- it is all too often a vehicle for infection by a virus. That is why DNA exchange is taboo in multicellular life. In computing, we are learning the importance of that taboo the "hard way" as we cope with increasingly dangerous digital viruses and worms. Polymorphic non-executable messages are far better suited to communication in multicellular systems.


Read more about Evolution of Computing

Last revised 5/30/2014