You've just cloned a unique cytokine named "zorf," then had the additional good fortune to publish your discovery in a top-tier journal. Your lab is the only source for this red-hot reagent. Within days of publishing the paper, you get a request for zorf from the Smith lab, which wants to test its effect on T cells. Publishing a paper in many journals obligates you to provide any unique reagents described in the article to researchers who request them. Such a requirement is entirely appropriate; without it, there might be no way to confirm a novel scientific finding using a difficult-to-obtain reagent.
You send the material to Smith straight away. A few days later, you get a new request for zorf from the Jones lab. Jones also proposes to test the effects of zorf on T cells. Sending zorf to the Jones lab could easily, in effect, pit the Smith lab's efforts against the Jones lab. You've never personally met either investigator, and now you find yourself standing on an ethical tightrope by potentially setting up a competition between these two groups. How much should you reveal to each party?
Say nothing? You could send zorf to the Jones lab without telling them that it has already been sent to another lab for what are ostensibly, as far as you know, the same experiments. If you take this approach, you risk setting yourself up for complaints and hard feelings down the line. Inevitably the Jones or Smith lab will publish their findings, leaving the other lab angry with you for not telling them that you had provided zorf to another lab for the same experiments.
Say everything? You could simply tell Jones that you have already sent zorf to Smith's lab, which plans to do exactly the same experiment that Jones has just proposed. The problem here is that there may be some type of confidentiality agreement in place with Smith (via a material transfer agreement) that prevents you from telling others about the experiments that he has already started. Even without such an agreement, Smith might be quite upset to learn from Jones at a conference that you told Jones about the experiments that Smith was doing with zorf.
As with most problems, the solution lies somewhere in the middle.
When you get the first request from Smith, tell Smith that while you can supply him with zorf, you cannot promise to do so exclusively. Thus, you have already warned him he may find out in the future that another lab is doing the same experiments as he is with zorf.
Also, try to get a more complete description of exactly what experiments Smith is proposing to do. A description on the order of "Test zorf on T cells" is simply inadequate. (Human or mouse? What type of T cells?) Then establish whether or not you are going to hold the details of Smith's experiments in confidence. (Smith may prefer to keep his details secret to avoid getting scooped. If you're collaborating with Smith, or want to establish an ongoing relationship with him, then it's in your interest to keep things secret, as well.)
If you get permission from Smith to share the details, then you can tell Jones when she approaches you that another lab (without identifying which one) has already requested zorf for similar experiments. Armed with this information, Jones may decide not to do the experiments that she was thinking of doing and instead point her lab's resources in a different direction. If you agree not to share details of Smith's work, this would not preclude you from telling Jones that you have already given zorf to another (unidentified) lab that is working in the same general area.
If you tell Jones that you've already given the reagent to another lab for the exact same experiments, and Jones says she still wants it, you should obviously give it to her. You've warned Smith and Jones that it's not an exclusive contract, and you've fulfilled your obligation to the journal to provide the reagent in response to legitimate requests. You should also consider the very real possibility that the first lab group might never complete its experiments.
Let's look at the other side of the collaboration coin. Suppose you need a collaborator to provide a critical piece of data to complete your draft manuscript on zorf. You know from experience that collaborations don't always generate useable data, for a variety of reasons. Would it be ethical for you to contact two labs and ask each of them to test zorf in the same experimental model, just to ensure that at least one of them generates good data?
This approach might be tempting but is fraught with problems. If both labs wind up doing the experiments, which one's results would you include in the manuscript? The second group won't be too happy with you, having generated the data you asked for, but then not seeing it published. Suppose the two groups generate different results? Which set do you choose? There's an easy way to avoid the drama: Just pick the collaborator that you think has the best chance to complete the experiments in a timely manner, and hope for the best.
By acting in an ethical manner in these various scenarios, you can only enhance (or at least not damage) the reputation of your own research group. Your standing in the scientific community as a fair collaborator can help expand your scientific horizons as other investigators seek you out to assist them with their studies.
Stewart Lyman is the manager of Lyman BioPharma Consulting in Seattle, Wash. Advice in this article is based upon his experience in his former position as director of extramural research at Immunex Corp. His Web site is www.lymanbiopharma.com