Replying to an editor after plagiarism accusation and manuscript rejection

Question

As the title of this question suggests, a manuscript that I submitted was flagged as plagiarized by iThenticate and was rejected by the editor.

The editor claims that the manuscript has 35% similarity with a previous manuscript that I submitted to the same journal. The email made me feel very bad, as the accusations are something that I have intentionally avoided throughout my publications. I also should note that I chose not to write repeated information and instead wrote a summary of some of the methods while citing the mentioned already published work.

Summary about the scope: it's mainly about sensors and machine learning, though both papers have very different goals and methodologies:

• Already published work: usage of many sensors to achieve two goals with machine learning regression and classification.
• Submitted manuscript: usage of one sensor to achieve a completely different goal from the already published work by machine learning classification and to explain the models with explainable machine learning.

We do not have access to iThenticate, but we do have access to Turnitin. The summary:

• Overall, there is 23% similarity with other published work.
• With the paper that the editor mentioned, there is only 8% similarity. The similarity mainly comes from affiliations, references, acknowledgments (funding agency and project number).
• The description of the sensor is flagged as well (name of sensor, company, location).

I have no idea how to do the following:

• To tell my mentors that this happened; I'm afraid that I will be in trouble if I do not sufficiently rebut the plagiarism accusation.
• To write to the editor to clarify what happened, though I suspect they already checked where the similarity is coming from.

Answer 1

1. The role of a mentor is to support you, not to punish you. So, IMHO you should/could have asked this very question to them. They presumably have some experience with this journal and can provide better advise then we can here (as we don't know the journal).
2. I suspect the editor has not carefully checked where the 35% similarity comes from, since your own analysis doesn't show undue similarities. I would write to the editor pointing our the similarities you are aware of and which are perfectly fine (affiliation, grant number, sensor make etc). If the editor rejects your arguments without really answering them, then I would consider a different editor or journal. There is no point in dealing with editors that see no sense.

Answer 2

Anecdotal answer here. I once had a similar experience, except it was 98% similarity...to the preprint of that paper on biorxiv. Never assume editors have put in the leg work, they have a lot on their plates and do it without pay. I politely pointed this out to them, I got an apology, it progressed and was published.