A “Quality-First” Credo for the Human Genome Project

The Human Genome Project is lurching toward large-scale genomic sequencing. Although we find the arguments in favor of this path compelling, there remain sobering uncertainties about the cost of producing genomic sequence on a gigabase-pair scale, the rate at which the needed sequencing capacity can be developed, and the extent to which compromises will be required in the quality of the final product. Of course, it is these very uncertainties that make the sequencing of the human genome a scientific and managerial challenge worthy of the committed attention of many hard-working and talented people. Policy decisions are now being made that will greatly affect how this talent is deployed during the years ahead. We argue here, on both scientific and managerial grounds, that it is essential that the Human Genome Project adopt a “quality-first” credo.

Scientific arguments for quality are rooted in a view of how the sequence will be used. The most common uses of reference sequences involve comparisons with other data. A vast number of sequences, derived from many sources—human and nonhuman—will be compared with the human reference sequence by future scientists. We should not prejudge either the comparison methods or the questions that the comparisons will address. Our goal should be to produce a reference sequence sufficiently good that comparisons made with it will rarely fail or produce misleading results …