This is not about how long it is ago. It is about that the mentioned co-author did not contribute to the current study. She only performed the sample acquisition 7 years ago. She did not even analyze the samples for the current study.

Whether the co-author has contributed to the current study, if the previous contribution is significant, there is no doubt.

even if the performance of that experiment is already published?

"Only" performed sample acquisition - so where would you be without those samples? If they were obtained last week would you feel differently?

If they were obtained last week, then the experiment itself would not be published yet. So authorship would be justified. At least thats the huge difference in my eyes.

In other words: credits were already given for the work done.

@Amanda No, the credit for the new manuscript have not been given yet, There is a joint paper by Gauss and Minkowski. Gaiss died before Minkowski was born, That is normal. Many joint papers by Erdos appeared after his death. That is also normal.

Your paper would not be possible without their contribution. There is no statute of limitations. So what if they got credit for some other paper as well? They did work that you needed to accomplish your paper.

In fact I would argue it would be quite inappropriate to exclude the author of part of your experimental process. That her experimental procedure was able to yield enough material for two separate scientific questions is irrelevant (and actually quite convenient for you).

@markvs How is this different from using other studies as a reference? Other results are referenced and credited all the time without giving authorship?

@Issel One important difference is the use of physical samples, which are in limited supply.

@mmeent Yeah I could see that, that does make a difference. Suppose a marine biologist gets lent a boat to do research on, would the captain/owner be listed as an author for supplying a hard to get and more expensive resource? It feels like you need to participate in the original research to be an author, and you are a contributor for providing resources.

@markvs "Many joint papers by Erdos appeared after his death. That is also normal": that is most certainly not remotely normal, at least not in the field of biology which is where the OP is from. It may be common in mathematics, but I can't think of a single such example in biology (or any other field I am familiar with). Also, this really isn't an answer, it's a glib (and unhelpful) comment. You have an answer in your comments, why not flesh this out so it can actually be helpful?

@terdon, the neuroscientist Pat Goldman-Rakic died in 2003, but is an author on 23 papers from her lab that came out between 2004-2008.

@terdon: You are trying to be rude. Bye!

@Matt I stand corrected, thank you. I still think calling it common is unjustified but fair enough.

@markvs I really am not. I honestly found this a bad answer. It isn't actually providing an answer, but is only a glib dismissal of the question. You explain it better in the comments, so I wish you would add some of your comments to flesh it out. As it stands, it comes across as rude and condescending. To me, anyway.

@markvs "Many joint papers by Erdos appeared after his death. That is also normal" xkcd.com/599

I downvoted - this answer makes no attempt to explain why published contributions would require co-authorship, which (if it is the intended argument) is patently absurd. Should Einstein be a co-author on every GR paper, because it uses his equations? Of course not. If there is a more subtle argument, it isn't apparent in the answer or the comments.

The co-author was a co-author of the previous paper which uses the same data. Einstein, etc. did not collect data for every paper which uses his equation. What is proposed by the OP is morally and academically wrong. You may pretend that you do not understand the difference.

@markvs its not the same data that is used. Its the same tissue sample. And this sample was analyzed 7 yrs ago and published. Now it's analyzed again for a different outcome and published.

@Amanda It depends on what you call "data", a sample is raw data. I guess, collecting it required some non-trivial effort. I am not a biologist, though. I am a mathematician.