Multicellular Computing: Biological Information Processing
Cells and computers process information quite differently. Yet computers are the basic unit of the digital world just as cells are the basic unit of life.
A National Science Foundation workshop report points out that, “ A series of discoveries over the past fifty years have illuminated the extraordinary capabilities of living cells to store and process information. We have learned that genes encoded digitally as nucleotide sequences serve as a kind of instruction manual for the chemical processes within the cell and constitute the hereditary information that is passed from parents to their offspring. Information storage and processing within the cell is more efficient by many orders of magnitude than electronic digital computation, with respect to both information density and energy consumption.” Computing professionals would do well to understand the parallels too.
All living organisms, from single cells in pond water to humans, survive by
constantly processing information about threats and opportunities in the
world around them. For example, single-cell E-coli bacteria have a
sophisticated chemical sensor patch on one end (orange molecules in image below) that detect several
different aspects of its environment. The sensor's output biases the cell's movement toward attractant
and away from repellent chemicals. Each individual sensor has a dynamic range of about 102, far less than the range encountered in their environment. However the sensor system is cleverly coupled such that, jointly, the sensors have a dynamic range more than 105.
Information processing machinery within single cells involves a complex network of tens or hundreds of thousands of protein mechanisms, genes and gene-expression control pathways that dynamically adapt the cell’s function to its environment.
We cannot directly compare the information processing “power” of a cell to that of a computer. Size, power usage, robustness, and computational density clearly favor the cell whereas repeatability, precision and linear computational speed clearly favor the computer. Cells use a highly parallel architecture whereas computers use a serial architecture. Cells exploit randomness whereas computers do everything possible to suppress it. And cells apply themselves to quite different tasks than computers do. Single cell organisms are self organizing, gather their own energy, reproduce themselves, and defend themselves. All of those tasks require substantial information processing, In contrast, computers rely on their human "masters" for most of those functions. So we know, or fancy that we know, the totality of the information processing done in a computer and are still largely ignorant of the information processing done in a cell.
We may attempt to deduce something about the information processing capacity of a cell from how much "code" it contains. A cell's DNA plays a role roughly analogous to program code in a computer. Although some of the DNA has been labeled as "junk," new research casts doubt on that notion, A human cell contains about 3.5 billion bases. Since each base is one of four possibilities (A, T, C, G), each encodes 2 bits of information. So the human genome contains about 7 billion bits, or roughly a Gbyte of "code." E-coli bacteria are relatively simple with about 5 million bases. Other single cell organisms can be much more complex. An Amoeba (one of the most complex single cell organisms) contains a hundred times more DNA than a human cell. This is one consequence of the fact that single cell organisms cannot specialize the way cells in multicellular organisms do.
At this point even the most powerful computer (IBM’s Blue Gene)
cannot simulate all the processes of a single cell in real-time. In fact, it
can’t even simulate one process – the folding of a single protein molecule
into its working shape – in anything like real-time. On the other hand, even
the most complex cell cannot accomplish what a mundane four function
calculator can do, especially in terms of precision and reproducibility.
Cells and computers simply face different tasks and have different capabilities.
Nonetheless, individual computers and single cells play similar roles in the large-scale sweep of evolution. Just as computers are the initial unit of computation, cells are the initial unit of life. And the challenges of communication and collaboration between networked computers are similar to those between cells in a multicellular organism.
Contact: sburbeck at mindspring.com
Last revised 1/6/2010