Discerning dangerous gain of function: most gain of function (GoF) research does not involve infectious microbes

The article reviews the use of the term "gain of function" (GoF) and "dangerous gain of function" (DGoF) in both the regulatory space relevant to the United States of America and the biomedical literature.

We identified nearly 20,100 papers in PubMed which mentioned "gain of function" in the title or abstract. Through a series of machine and human evaluations, these were resolved to 145 that appeared to be dangerous gain of function (DGoF). In addition, we evaluated some papers that did not mention gain of function but involved expression of recombinant sequences of concern (SoCs) in heterologous microbes where the latter is endowed with pathogenic or toxic function as a result of the transfer. For 22 viral GoF papers, we discuss criteria leading to determinations of whether DGoF occurred with reference to the seven DURC criteria and discuss challenges in evaluating them.

We found that the overwhelming majority of the use of the "gain of function" term is not concerned with dangerous gain of function research. Less than 1% of the articles that use the GoF term discuss DGoF research. Over 40% of those are review or policy papers that do not deal with empirical research. Of the remaining experimental GoF articles, the majority involve antimicrobial resistance (AMR) in pathogens. In addition, it appears that most research that could be considered as DGoF according to the seven criteria for dual use research of concern (DURC) is not characterized by the authors as GoF.

1) most GoF research is not DGoF, 2) most DGoF is not referenced with the GoF term, and 3) discerning DGoF from GoF is not always straightforward but is helped by recognition rubrics, particularly the DURC criteria, as well as examples of SoCs with written functional definitions.