Planned alterations to native chromosome sequence are potentially infinite in number and so a great deal of thought must be given to the specific alterations to be incorporated into the synthetic chromosome(s). In general, we are making modest changes predicted to have minimal or no impact on fitness of the organism under laboratory conditions.

These changes are introduced in a stepwise, “bottom up” approach. The advantage of a stepwise iterative approach is that the “investment” made in the project at any given stage is limited to about 30kb, so if a particular segment is incompatible with viability, or gives rise to a slow-growth phenotype, the segment can be re-synthesized after the nature of the growth defect is mapped and diagnosed (by singling transforming in the 10kb pieces, or by mapping which segments of a given 30 kb chunk are incorporated during transformation [PCRTags]. This very practical strategy is in contrast to a “home-run” strategy in which a more aggressive design is conceived, a lot of work invested, and then one hopes the outcome is a viable cell. A key design principle is that we will not attempt to make individual changes likely to result in a decrease in fitness. The cumulative effect of such changes is likely to result in an unviable yeast cell within a relatively low number of cycles. Nevertheless, we are building in a system to allow subsequent evolution of a minimal genome, as well as faster growing variants and variants with a potentially infinite variety of biological adaptations by building into the genome a “SCRaMbLE” controllable genome evolution system that can be controlled by the experimentalist at will.