Imitation of Life
Can a computer program reproduce everything that happens inside a living cell?
Almost 30 years ago, Harold J. Morowitz, who was then at Yale, set forth a bold plan for molecular biology. He outlined a campaign to study one of the smallest single-celled organisms, a bacterium of the genus Mycoplasma. The first step would be to decipher its complete genetic sequence, which in turn would reveal the amino acid sequences of all the proteins in the cell. In the 1980s reading an entire genome was not the routine task it is today, but Morowitz argued that the analysis should be possible if the genome was small enough. He calculated the information content of mycoplasma DNA to be about 160,000 bits, then added:
Alternatively, this much DNA will code for about 600 proteins—which suggests that the logic of life can be written in 600 steps. Completely understanding the operations of a prokaryotic cell is a visualizable concept, one that is within the range of the possible.
There was one more intriguing element to Morowitz’s plan:
At 600 steps, a computer model is feasible, and every experiment that can be carried out in the laboratory can also be carried out on the computer. The extent to which these match measures the completeness of the paradigm of molecular biology.
Looking back on these proposals from the modern era of industrial-scale genomics and proteomics, there’s no doubt that Morowitz was right about the feasibility of collecting sequence data. On the other hand, the challenges of writing down “the logic of life” in 600 steps and “completely understanding” a living cell still look fairly daunting. And what about the computer program that would simulate a living cell well enough to match experiments carried out on real organisms?
As it happens, a computer program with exactly that goal was published last summer by Markus W. Covert of Stanford University and eight coworkers. The program, called the WholeCell simulation, describes the full life cycle of Mycoplasma genitalium, a bacterium from the genus that Morowitz had suggested. Included in the model are all the major processes of life: transcription of DNA into RNA, translation of RNA into protein, metabolism of nutrients to produce energy and structural constituents, replication of the genome, and ultimately reproduction by cell fission. The outputs of the simulation do seem to match experimental results. So the question has to be faced: Are we on the threshold of “completing” molecular biology?