In early 1989, the American Journal of Clinical Nutrition was preparing to publish a paper on diabetes in nursing mothers when its editors received a short letter. "Dear Sirs," the note began, "Dr. Lois M. Freed, one of the authors of manuscript #1238RR, has requested that her name be removed from authorship. In accordance with her wishes, please delete her name."

There was no further explanation in the letter, no clue as to the deep disagreements that lay behind it: Freed, a researcher at Georgetown University Medical Center, was convinced that her co-authors had omitted significant data that, in her view, affected some of the paper's conclusions. Freed had urged that the paper be changed or withdrawn. The lead author of the paper disagreed with her, and the article was published without Freed's name.

Nearly two years later, the conflict has mushroomed. The chief researcher, Georgetown scientist Margit Hamosh, has been forced to defend herself against charges of falsifying data, plagiarism, and mismanaging research in her laboratory. Freed, who says she was dismissed from her job after the incident and is now a researcher at the National Institutes of Health, has filed a lawsuit against Hamosh and Georgetown. NIH, which funds the bulk of Hamosh's research, has launched its own probe of the allegations against Hamosh.

Further, one member of Georgetown's committee set up to investigate complaints of research fraud has resigned from the panel in protest over the university's dismissal of the case without a full investigation. "Scientists can make honest errors, can be sloppy and negligent, but when there are charges of willful deceit, these must be investigated fairly, objectively and impartially . . . The integrity of science was the major victim," Dinkar K. Kasbekar wrote in a confidential May 11, 1989, letter to NIH, according to officials who have seen it.

The case has left those involved bruised, angry and defensive. Hamosh has rejected the allegations as "frivolous" and "pure fiction" in a written response to university officials. Georgetown defends the committee's action as proper, says it is cooperating fully with the NIH probe and also denies any wrongdoing in Freed's lawsuit.

But whatever the outcome of the NIH investigation, the episode illustrates how difficult it is for universities to make sure that allegations of scientific misconduct are adequately investigated.

Just two years ago, after a series of celebrated cases, there were demands from Congress and elsewhere for better protection for the $6 billion in research funds dispensed annually by the federal government. Some of the strongest attacks came from Rep. John D. Dingell (D.-Mich.), the chairman of the House Energy and Commerce Committee. At one hearing, Dingell blasted the scientific community for its "apparent unwillingness . . . to deal promptly and effectively with allegations of misconduct."

NIH got the message. "It became clear that if the scientific community did not get its act together, with NIH as its leader, it was going to be done to us, and by another entity," said Suzanne Hadley, deputy director of the federal Office of Scientific Integrity at NIH.

That office was created in 1989 to oversee the new system that placed the initial burden on universities themselves to investigate fraud -- fabricating data or knowingly publishing false information. The new system also requires institutions seeking federal research funds to formalize procedures for investigating fraud, or risk losing their grants.

Locally, Georgetown took the lead, acting even before the new rules went into effect. The medical center appointed a standing fraud committee in late 1988 -- with Hamosh as one of the original members. Medical School Dean Milton Corn discussed the university's goals in a December 1988 faculty newsletter: "Fraud has no place in any form of scholarship," he declared. "The academic community must avoid even the appearance of tolerance of scientific misconduct."A Leader in Nutrition

The stakes are high for Hamosh and Georgetown. Hamosh, 57, chief of the Division of Developmental Biology and Nutrition, is a prolific and respected pediatrics researcher who has brought Georgetown almost $4 million in research grants over the past 15 years. She has built a career around her work in infant nutrition and is a nationally recognized expert in the field. The grants she brings Georgetown could be terminated if NIH finds that any misconduct occurred.

In a brief statement to The Washington Post, she said she could not address the specifics of the allegations because the matter is confidential under university and NIH guidelines.

Hamosh came in 1974 from NIH where, along with scientist Robert Scow, she discovered lingual lipase, an enzyme that aids in the digestion of milk. It became her specialty. "She's literally the tops in this particular area," said Robert G. Jensen, professor of nutrition at the University of Connecticut. "She has, in the last decade, done practically all the important work."

All told, Hamosh has published 90 original papers, 63 book chapters and reviews, 180 research abstracts and a textbook. In 1988, she received the Dean's Prize for Excellence in Research; she has sat on important committees of the National Academy of Sciences and NIH, advises a leading journal and helped to found an international research society on nutrition for nursing mothers.

Her work has influenced the development of infant formulas and led to a greater understanding of the benefits of breast-feeding, experts said. Her peers say she is a person of high integrity who shares credit generously. "I have a sense that Margit has always upheld those ethical goals," said Cutberto Garza, director of nutritional sciences at Cornell.

At Georgetown, she consistently has ranked high on the list of NIH grant recipients, records show. The university has received 52 percent of her NIH grant money for indirect costs of her work, and she has her own lab and research staff. She also works closely with her husband, Paul, a physician on the faculty.

After more than 15 years at the university, Hamosh and her husband are firmly entrenched in Georgetown's academic community -- but they have also ruffled the feathers of some colleagues. "The Hamoshes are themselves very strong and controversial figures at Georgetown," said one Georgetown official. "They, particularly, are people who express their opinions very strongly, don't concede much in arguments . . . and have over the years, in addition to having done a lot of very fine scientific work, probably irritated a large number of people."

Hamosh, in a short statement to The Post,said, "I am confident that my peers, who are familiar with my scientific work over the past three decades, know that I have always acted properly and in accordance with accepted standards of scientific inquiry." A Disillusioned Researcher

Freed, 39, said she admired Hamosh when she first went to work for her in 1982. Under Hamosh, she prospered: Freed's dissertation won the 1986 Young Investigator Award from the American College of Nutrition as the outstanding PhD thesis in the field of nutrition. In 1987, after a strong recommendation from Hamosh, Freed was appointed to the faculty and eventually became a research instructor.

"Dr. Freed has a sharp mind, superb technique and great intellectual curiosity, the hallmarks of a fine investigator," Hamosh wrote in her recommendation at the time. "She is also an outstanding teacher, able to convey complex ideas in an easily comprehensible way . . . She will be a valuable asset to our Department."

Yet, Freed said, she began to see things that disturbed her in the lab and heard from other researchers who felt the same way.

"It just always seemed like there were little compromises on little things," said Carla M. York, who worked as a research assistant for Hamosh for three years and whose name appears on three Hamosh papers. "It wasn't like the ultimate outcome was scientific knowledge. It was like, just publish whatever you can -- to get more grants and more money."

"A lot of this was simmering in the lab for a long time," said Susan E. Berkow, another former Hamosh researcher and co-author of five papers.

In January 1989, Freed found what she said were discrepancies in the draft paper for The American Journal of Clinical Nutrition that led her to file her complaint with the university.

The paper suggested that in nursing mothers, diabetes could change the way the mammary gland metabolizes fat -- the chief source of calories for a nursing baby -- and thereby affect the composition of milk being given the baby. Freed noticed that a key diagram, based on research she had done several years earlier, did not reflect some data she had collected; this omission altered the shape of the curve. Freed said she told Hamosh: "This doesn't look like what I generated."

At first, Freed said, she thought it would not affect the paper's conclusions, but she later found that other data, which might have pointed to an alternative explanation for the paper's findings, also had been omitted. "Simply correcting the graph," she wrote to another senior author, "is not sufficient to make the manuscript acceptable for publication."

Hamosh redid the graph but declined to make the broader changes, saying Freed was wrong in her assessment of the data. What happened next is disputed: Freed claimed she was dismissed for challenging the article; Hamosh has said she was merely accepting a resignation Freed had earlier offered.

In her lawsuit, Freed contests her dismissal. Last week, Freed's lawyer, Lynne Bernabei, said she thought that in the course of the case Georgetown was trying to hide information from the government. Before the trial, the university has asked a D.C. Superior Court judge to bar Freed from sharing any information obtained during the discovery phase with NIH investigators. Georgetown Fraud Panel

After the disagreement between Hamosh and Freed over the content of the article, Freed filed a formal three-page complaint with Dean Corn and the university's Scientific Fraud Committee. It laid out approximately six allegations of misconduct. Two months later in March 1989, the panel took up the complaints against Hamosh. The allegations had by then broadened beyond the scientific paper: Hamosh was being accused by another researcher of plagiarism and by Freed of claiming in published work and federal grant proposals to have successfully carried out experiments that either had not worked or had not been done.

Hamosh denied any plagiarism, saying she had properly credited other scientists for their work and that her laboratory had completed and reported all experiments.

The committee consisted of about ten medical and science professors and was chaired by Robert C. Baumiller, chief of genetics in the department of obstetrics and gynecology. Hamosh, one of the members, disqualified herself from the case. Kasbekar, professor of physiology and biophysics at Georgetown for 21 years, was one of two other professors who did not participate in the committee deliberations, because he had once co-authored three articles with Hamosh.

Before the committee acted, Freed says, she was summoned by Baumiller, who attempted to dissuade her from pursuing the complaint. "Rev. Baumiller told me that I needed to be aware that Dr. Hamosh was a member of the Scientific Fraud Committee, that the members were her friends and would 'rally around her,' " Freed wrote in her account to NIH.

"Rev. Baumiller asked me if I fully understood the potential outcome of a fraud investigation -- that if evidence of fraud were found that the University might be forced to return grant money." Baumiller declined to respond to Freed's account.

Freed said she refused to back down. Baumiller then circulated to the full committee the complaints of Freed and Laura L. Richardson, another former researcher in Hamosh's lab. Each also appeared before the panel, as did Hamosh.

After reviewing the allegations against Hamosh in strict confidence, the committee dismissed the case in mid-March without conducting a full investigation. As a result, Georgetown did not have to report the allegations to NIH or to any journal where the data in question may have appeared.

"After serious review," Baumiller wrote to Georgetown Vice President John F. Griffith, "the Committee's consensus was that a formal investigation was not warranted, and no further actions were necessary." Baumiller also noted that "those bringing the allegations presented their concerns in a dignified and respectable manner . . . The committee found no reason to find that those bringing their concerns were bringing them with malicious intent." A Question of Favoritism

But the affair did not end here -- in fact, it was just beginning. Kasbekar's letter to NIH characterized the committee's handling of the case as "highly improper and prejudicial." Kasbekar, reached by phone, confirmed his resignation and that he had written to NIH but said that he could not comment because of the school's policy of confidentiality.

To Freed and several other former Hamosh researchers and lab technicians, the committee members were trying to protect a colleague rather than investigate the charges fully. "She's one of them. She's been their professor for a long time . . . She's making a lot of money for them," said Sania Amr, a physician in Baltimore. Amr, a former assistant professor of pediatrics at Georgetown, collaborated with Hamosh on two papers during 1988 and 1989.

NIH opened its own investigation into Hamosh's laboratory practices after receiving Kasbekar's letter and hearing from Freed. As part of its lengthy probe, it has convened a panel of outside experts and has taken testimony from Freed and several other former Hamosh researchers.

Georgetown officials strongly defend their handling of the case. "Georgetown University has full confidence in the performance and procedures of its committee," the university said in a brief statement. The committee "discharged its responsibilities in a diligent and professional manner."

University officials noted that Hamosh's NIH grants in the fiscal year 1988-89 accounted for less than 2 percent of the university's $36 million NIH total.

Baumiller also rejected any suggestion that the panel was looking to protect Hamosh or her federal grants.

"You've got a bunch of academics, and they want to do what's right," he said. "The committee did what it considered to be correct, and labored intensively over each of the issues . . . I feel very comfortable with that, and it's out of my hands. I haven't lost a bit of sleep over it." Unresolved Issues

Defenders of the Georgetown committee's handling of the Hamosh case said there may have been a lack of understanding or agreement among some members over some issues -- such as whether allegations of misconduct occurring in unpublished, confidential grant proposals to the federal government deserve any less scrutiny than published data in scientific journals. They also pointed out that NIH's final rules on investigating misconduct had not yet been issued when the Hamosh case arose and that committee members, facing their first case, lacked experience.

They also said Georgetown has since broadened the committee's membership and expanded its jurisdiction to other parts of the university.

Others said the Hamosh case illustrates the possibility that the new system demands too much of a university. They ask: Can professors be expected to dispassionately investigate a colleague? Have universities been placed in an impossible position by being asked to rule on allegations of misconduct, particularly when federal grant money is at stake?

Corn, who left Georgetown in July 1989 to take a senior post at NIH, said the Hamosh case was a first test of the new system. Corn was not a member of the fraud panel and played no role in the case, but he has given much thought to misconduct in science.

"Publication pressure, sensible or not, exists, and we should view with some sympathy the occasionally encountered colleague who neglects almost all other responsibilities in the zeal to churn out papers," wrote Corn to the faculty in 1988. "Zeal is different from fraud, and for those few who lose their way by confusing scientific reports with the writing of fiction, there can be little sympathy."

The integrity of data is critical in the research community, where the progress of science is seen as a chain with each published paper a key link. While science is, of course, largely the process of correcting earlier mistakes, inadequate judgments and erroneous assumptions, fraud is damaging "because it undermines the whole scientific process upon which the welfare of our society depends," said Arnold Relman, a physician and editor of the New England Journal of Medicine.

"Science is one of those activities," said Relman, "which rest to a very considerable degree on trust and integrity and self-correction and responsibility."

In investigating allegations of fraud by peers, scientists are confronted not just with the complexities of research but also the passions of different personalities. Universities, moreover, are generally reluctant to publicize their shortcomings.

"As long as they {the members of Georgetown's Scientific Fraud Committee} really thought that there was not enough here to justify a full investigation, it was in the interest of the university to close it at the inquiry stage, because that kept it on Reservoir Road" -- Georgetown's street address, said Corn, referring to the rule that NIH does not have to be notified unless a full investigation is undertaken.

Corn said he felt "regret" when he heard the case had been closed at the inquiry stage. A full probe would have left no doubts about Hamosh's integrity or about the process, he said.

The Hamosh case "tested the system, and it tested the fact that an apparent procedure developed by a committee . . . creaked and strained when actually faced with something," he said.

Staff writer Susan Okie and researcher Lucy Shackelford contributed to this report.