The University has required a post-doctoral student in biology accused of research fraud in the laboratories of the Nobel Prize winning Loeb University Professor Walter Gilbert '53 to leave Harvard for one year.

Dr. Carl E. Fulwiler, a research associate in biology, confirmed yesterday that he would leaving Harvard but offered no further comment.

Fulwiler was one of about six researchers in Gilbert's lab studying zebrafish as a model for early development of genes. Fulwiler allegedly changed data in his laboratory notebook to make it appear as if an experiment with the fish's genes had worked better than it actually had.

Gilbert the data in the draft of a paper that was reportedly going to bepublished in the journal Nature, accordingto lab sources. Sources said a post-doctoralstudent in Gilbert's lab discovered the errorbefore publication.

Gilbert did not return a phone call yesterday.

No wrongdoing has been attributed to Gilbert, aformer chair of the Department of Cellular andDevelopmental Biology.

But Gilbert's laboratory has nevertheless beentorn apart by the allegations, lab sources said.The University's decision to imposes sanctions onFulwiler will likely make the atmosphere atGilbert's lab even more tense, they added.

Sources said that the case was referred to thefederal Office of Research Integrity (ORI) inBethesda, Md.

ORI would not comment on the case because of apolicy which prevents them from discussing opencases and cases that have been closed withoutfinding evidence of scientific misconduct, saidDr. Kay Fields '63, a senior scientist at ORI.

The office works to "protect the public money"by imposing financial sanctions on researchers whofalsify evidence. Researchers found guilty offraud can be banned from receiving federal fundsfor up to ten years. Fields said.

An ORI investigation of a case usually takesabout a year, Fields said.

Universities can impose their own sanctions onresearchers whether or not they are found guiltyby ORI "as long as they are vaguely appropriate,"Fields said. Sanctions can range from asking themto leave a lab to requiring them to seekcounseling.

Universities often act more quickly than ORI."It can announce its findings before they are sentto accepted by us," Fields said.

Big-name professors like Gilbert, who oftenhave large labs and little free time to policethem, may be particularly susceptible to researchfraud by lab workers.

But some professors said that Gilbert takescare to avoid fraud better than most.

For instance, a two-page instruction sheet onrecord keeping is posted in Gilbert's lab, tellingresearchers to write everything down and keep neatnotebooks