A new law to require federal licensing of laboratories that combine genetic material to make novel forms of life was urged yesterday by a committee representating 16 federal agencies.

If passes such a law would for the first time let the government dictate what kind of basic research a scientist could and could not do.

The committee, according to government sources, recommended that the Secretary of Health, Education and Welfare regulate and inspect public and commercial laboratories doing such work and that federal law supercede any local rules like those last month passed in Cambridge, Mass.

The committee recommendation, made in closed session, will be sent to HEW Secretary Joseph A. Califano Jr. today.

Scientists and committee members have worried about what many call the "awesome" preceden of letting the government decide wha a scientist may or may not do.

Laws now regulate research involving human beings, and they require laboratories to obey safety standards. Scientists must be licensed by the Nuclear Regulation Commission to use radioactive isotopes. The government heavily influences research by deciding what it will fund.

But no law other than general safety standards has ever governed scientists working in so esoteric a field as combining the DNA - deoxynbonucleic acid, the basic genetic material of different organisms.

Scientiest themselves first called attention to the potential risks in joining DNA from different creatures - and, specifically, in inserting mammal genes and other genes into colonies of bacteria.

They agreed to avoid experiments that might create new diseases or adversely affect human, animal or plant genetics.

Last June, at scientiest' proding, the National Institutes of Health issued safety guidelines for scientists getting federal funds.

In September, former President Ford named the interagency group to decide whether the guidelines should become law and should also apply to private firms. He named representatives from the White House, HEW, Agriculture, Defense, Commerce, Labor, Interior, Justice, State, Transportation and energy, environmental and science agencies, with Dr. Donald Fredrickson, NIH director, as chairman.

The committee concluded that Congress in general should:

Request registry, licensing and inspection of all laboratories where the so-called "recombinant DNA" research is conducted.

Apply federal rather than state or local laws in a city like Cambridge, which has already forbidden some types of this research. Scientists fear that a patchwork of local laws will retard progress.

Provide a mechanism to let private firms, for example, drugmakers, keep some secret until they apply for patents, although the HEW Secretary could determine what projects are eligible.

House Health Subcommittee Chairman Paul Rogers (D-Fla.) last week introduced a bill to achieve similar goals. But he called it a mere "vehicle for discussion" as his subcommittee and others consider this and other legislation.

The Rogers subcommittee starts three days of hearings this morning. More are due in April before the Senate Health Subcommittee headed by Edward M. Kennedy (D-Mass.).