Scientists at the California Institute of Technology announced today that they have developed the first automated machine to analyze the structure of DNA, the body's basic genetic material. The device is expected to accelerate the search for cures for cancer and genetic diseases such as cystic fibrosis.

Caltech and National Science Foundation officials said they expect the $90,000 DNA "sequencers" to be purchased by every major medical research laboratory in the country and eventually to provide data for use in hospital diagnoses.

Leroy E. Hood, 47, Caltech's biology division chairman and head of the team that invented the device, said it will have immediate impact on research into diseases such as cystic fibrosis and hemophilia, in which scientists painstakingly attempted to isolate abnormal genes.

"This machine will automate many of the tasks involved in DNA sequencing now done laboriously by hand," Hood said. "It will allow researchers to analyze the structure of DNA molecules far more quickly and less expensively than ever before."

Steve Tronick, who conducts research on gene structures and cancer at the National Cancer Institute, said the device should "save tremendous amounts of time" in an area "extremely important to cancer research and all biology." He said he had not read the technical report on the invention in the June 12 issue of Nature but expects the device to be a success.

DNA, short for deoxyribonucleic acid, is a long, double-stranded molecule that carries the blueprint for the construction of living organisms. DNA molecules make up the 23 pairs of chromosomes in each human cell. Aligned single-file on each chromosome are genes that determine individuals' height, hair color, foot size and everything else that makes them human and unique.

The Caltech device uses a laser, dyes and a microcomputer to exploit the fact that each DNA molecule resembles a string of about 1,000 tiny beads. Each "bead" is one of four basic units called nucleotides, either adenosine, guanosine, cytosine or thymidine. Scientists call the nucleotides A, G, C or T.

The sequence of the nucleotides along the molecule determines its genetic information. The machine can determine that sequence 100 times more efficiently than a human researcher, accelerating the discovery of the single misplaced nucleotide that often identifies a cancer-causing or disease-prone organism.

Until now, researchers have used a manual method of DNA sequencing developed by Nobel laureate Fred Sanger of Cambridge University.

Jane Z. Sanders, a biologist on the Caltech team, said Caltech senior research fellow Lloyd M. Smith developed a way to color code the ends of fragments of DNA, green for those ending with the A nucleotide, yellow-green for G, orange for C and red for T.

Working with grants from the National Science Foundation, the Baxter Foundation, the Weingart Foundation, Monsanto Corp. and Upjohn Inc., the Caltech team injected color-coded DNA fragments into an ultra-thin glass tube full of gel and forced them down the tube with an electrical field. The smaller fragments travel faster than the larger, and they quickly arrange themselves in order of length, each fragment one nucleotide shorter than the one above it.

A laser beam then catches each fragment as it goes by, helping to illuminate and identify the color code of each fragment end. The signal is transmitted to a computer, which produces a long string of As, Gs, Cs and Ts denoting the sequence of that molecule. That sequence then can be compared to the proper sequence for that organism.

Hood said he expects at least 1,000 of the devices to be sold soon after a Foster City, Calif., firm begins marketing them in about nine months. Caltech officials declined to say how much the university will receive in royalties. Hood's team also included Robert J. Kaiser, Peter Hughes, Christopher Dodd and Stephen B.H. Kent.