||neither of whom are wall-flowers. Near-slanderous public attacks from both sides in the press led to the forced resignation of Watson as Genome Director and the resignation of Healy under duress. Under new leadership, N1H backed off patenting and the issue has now reached a seemingly sensible position where the utility of individual sequences needs to be demonstrated and the patent depends on demonstrated uses. However, tests of such patents have yet to play out in court. Only a small fraction of the TIGR cDNA sequences have made it into the database. However, Merck has supported Bob Waterston to sequence a few hundred thousand cDNAs and deposit the data directly in the public data base. This contribution has been very useful to the research community. One suspects that this decision by Merck was not totally altruistic, however, as it did have the effect of minimizing the value of TIGR's resource. Future of Sequencing DOE and NIH have agreed on a new sequencing goal: to reach a capacity of 50 megabases per year by the end of the next 5 years. That figure seems too conservative. The goal ought to be a megabase a day in a laboratory of modest size. It can be done. The trick will be to build a system that deals with the whole flow of the process from chromosome fragments in one end to sequence out the other, with well tuned rates for each process and the right level of automation. At our Utah Center for Human Genome Research, we have set about building such a system. We have a fertile mix of 40 engineers, molecular biologists and computer scientists developing a sequencing "pipeline". The DNA flowing in has been ordered and organized using biological tricks so that the high redundancy required by the shotgun approach is avoided. These samples are processed and analyzed in complex mixtures and the individual sequences are deconvolved in an automated manner, with an output of digital images of sequence ladders that are electronically read into the A, G, C and T sequences. The core technology is a new instrument that accepts a nylon membrane with the complex patterns of mixed samples and performs unattended serial rounds of probing, scanning and reading to give the individual sequences. The output will near 100,000 bases per day, per machine, and the machines can run unattended for many days. The instrumentation and the balanced flow save labor and increase reliability. Trained technical expertise then goes into decision-making and trouble-shooting, where humans are good. A modest-size team can run many machines. With this system, reaching 1 megabase per day is not an unreasonable goal.