The National Academy of Sciences has endorsed a controversial $3 billion program to decipher all of the genes in the human body, according to a report released yesterday. The project has generated widespread debate in the biomedical research community.

Battle for control of the project has begun between the National Institutes of Health (NIH) and the Energy Department (DOE), which has taken the initiative and proposed to move the so-called genome project into two of its national laboratories.

The academy does not determine who should run the research. To produce the study, it has asked Congress to approve an additional $200 million annually for 15 years for biomedical research, a 3 percent increase.

Congress has been considering the genome project and, in December, appropriated $17.4 million to the NIH and $10 million to the DOE to begin technology development and pilot projects this year.

Completion of a map of all human genes would be "a fundamental advance" with benefits in many areas of medicine and biology, said Dr. Bruce M. Alberts of the University of California at San Francisco, chairman of the academy's National Resource Council committee, which issued the report recommending the project.

Such a map would advance scientists' understanding of how specific gene defects produce diseases such as cancer and would aid diagnosis and treatment of such inherited disorders as cystic fibrosis and familial Alzheimer's disease, said Dr. Victor A. McKusick, professor of medical genetics at Johns Hopkins University Medical School, another panelist.

McKusick said it could also lead to development of tests to predict whether an individual is genetically predisposed to developing such conditions as cancer, heart disease, hypertension or diabetes.

"We needed to make a statement" about the importance of this research "just as we did with AIDS," Alberts said.

Each of the 10 trillion cells in the human body contains about 100,000 individual genes that determine such physical characteristics as height or eye color. Genes are encoded by the chemical deoxyribonucleic acid, DNA, which in turn has about 3 billion subunits.

The proposed project would determine the identity, position and function of the 100,000 genes and eventually determine the order of the subunits.

Since each human differs from another human by about one of every 1,000 subunits and since DNA samples used to produce the map would come from many human donors, the result would be a "representative map" rather than a record of any individual's genes, Alberts said.

"We're talking about a genetic data base like a dictionary" to which scientists could refer, he said.

To manage such a huge project, the academy's committee said it envisions a centralized facility to provide genetic material to be analyzed and to manage and make available the incoming data.

Individual research labs would compete to perform the analysis. A scientific advisory board would be appointed to help determine that the individual projects were done properly and produced useful data.

Proponents have said the huge amount of genetic information gained from the research is expected to speed understanding of basic human biology and evolution, since the genetic makeup of animals also is to be analyzed for comparison with human genes.

Opponents in the biological community have expressed concern that a huge project would divert funding from other equally important research. The academy specifically recommended new appropriations to support this research as a way of overcoming this objection.

The Energy Department originally became involved in studying chromosome damage as a means of understanding the effects of radiation upon the body. DOE laboratories at Los Alamos, N.M., and Livermore, Calif., are conducting gene research.

In addition, a computer at Los Alamos is the repository of Genbank, a data bank that cooperates with a similar facility in West Germany to record findings of gene sequences as they are published.