Hello, Thank you for organising DREAM-Q. I am also taking part in the parallel DREAM-CACHE Target 2035 challenge, and I have read all eleven wiki pages of this project, so I am only asking about what does not appear there. **1. What earns authorship in DREAM-Q?** The DREAM-CACHE challenge states that participants who perform better than random selection are invited as co-authors. I could not find an equivalent statement for DREAM-Q. Is there a defined criterion (for example a rank, a threshold, or a comparison to random selection), or does the general DREAM Official Rule 9 apply as written? **2. Does Phase 1 alone count?** If a team completes Phase 1 (simulators) with a final submission and write-up, but is not able to run Phase 2 on quantum hardware, is that team still eligible for authorship or consortium listing? **3. Overlap with the DREAM-CACHE splits.** Both challenges draw on the same 400K ASMS library. Will the ~300 Phase 1 held-out compounds be drawn from compounds whose labels are released, or will be released, in DREAM-CACHE? Is using knowledge gained in DREAM-CACHE permitted here? **4. Two small inconsistencies on the pages, in case they are useful:** - The Phase 1 end date reads **27 November 2026** on the Phase 1 page and **11 December 2026** on the Timeline and Phases pages. Which is correct? - The "Register team" button on the Submit page uses challenge ID 4861, which belongs to DREAM-CACHE rather than DREAM-Q (4869). Thanks very much, Arnav Agarwal

Created by Arnav Agarwal ArnavAgarwal
Thank you for the clear answers, and for fixing the date and the Register button. One follow-up on the relationship between the two challenges. Your reply covers using DREAM-CACHE knowledge in DREAM-Q. We would like to check the reverse direction before touching anything: The DREAM-Q training data are labels on the same 400K ASMS library that DREAM-CACHE evaluates on. The DREAM-Q Data Use Restrictions say the data may be used only for participating in DREAM-Q. Are teams that take part in both challenges allowed to use the DREAM-Q training labels when building their DREAM-CACHE October (Active Learning Step) submissions? If yes, should that be declared as external_data_used in the DREAM-CACHE write-up? We will not use any DREAM-Q data for DREAM-CACHE until you confirm. Best, Arnav
Hello Arnav, Thank you for reading the challenge materials so carefully and for flagging the inconsistencies. Our responses are below. 1. Authorship Invitations to join the challenge overview publication as co-authors will be determined by the review panel and/or organizers. Decisions will consider both challenge performance and the quality and scientific contribution of the submitted methodology, rather than relying on a single predefined rank or threshold. 2. Phase 1 participation Phase 1 participation does not automatically guarantee authorship. Teams that complete Phase 1 but are unable to complete Phase 2 may be considered for authorship only on an exception basis, at the discretion of the review panel and/or organizers (based on performance, scientific contribution and/or other relevant factors) 3. Relationship to DREAM-CACHE The approximately 300 Phase 1 held-out compounds will be selected from compounds whose labels have not yet been released when the DREAM-Q dataset is prepared. Teams may apply knowledge and methodological experience gained through DREAM-CACHE to DREAM-Q, provided they comply with the relevant data-use terms and challenge rules. A key DREAM-Q requirement remains that the submission include an essential quantum component, together with evidence that its inclusion was important to the predictions. 4. Corrections Thank you for catching both issues: - The correct Phase 1 end date is December 11, 2026. We missed one page when updating the schedule, and this has now been corrected. - The Register Team button has been fixed to use the DREAM-Q challenge ID, 4869. Thank you again for the helpful questions and careful review. Best, The DREAM-Q Organizing Team

Authorship criteria, and whether Phase 1 alone counts page is loading…