Ethics

Ethics & Compliance Reviewer

SCORE: 5 CONFIDENCE: 5

Summary

This manuscript presents dnoise, a tool for removing uninformative points from native Bruker timsTOF mass spectrometry data while preserving analytical results. The work is ethically sound and compliant. No human subjects, animal research, clinical trials, dual-use risks, or restricted-consent data are involved. The work is a computational analysis of publicly deposited commercial standards (PXD070049). Funding is disclosed (NIH grants and Skaggs Graduate School support). Competing interests are explicitly declared as none. All required compliance statements are present and properly attributed.

Compliance Assessment

Human subjects / identifiable data: Not applicable. The benchmark uses a commercial three-species digest standard (human/yeast/E. coli mixture) with no human biological samples, no patient data, and no identifiable information.

Clinical trials: Not applicable.

Animal research: Not applicable. The manuscript explicitly states "involved no human subjects or animal procedures."

Dual-use / biosafety: Not applicable. The work is a data-processing tool with no biosafety implications.

Restricted-consent data: Not applicable. All data are from a public PRIDE repository (PXD070049) with no access restrictions.

Funding disclosure: Present and complete. Five NIH grants are named with their recipients (R01 HL165168, R01 AG077046, R01 MH100175, R01 AG075862 to J.R.Y.; R01 MH132570 to L. Ye; U01 AG088679 to S. A. Lipton), plus Skaggs Graduate School support to P.T.G.

Competing interests: Explicitly declared as none: "The authors declare no competing financial interest."

AI disclosure: Present and transparent. The authors disclose use of Anthropic Claude for manuscript drafting, editing, and software-development assistance, with the explicit statement that "no AI tool directly generated an image or produced or altered experimental data" and that "the authors reviewed and edited all content and take full responsibility."

Data and code availability: Complete. Raw data are in PRIDE (PXD070049), dnoise is open-source (MIT licensed) on GitHub and crates.io with a versioned Zenodo archive (v0.1.0, doi.org/10.5281/zenodo.21959649). Parameters, configurations, and reproduction details are in the manuscript and Supporting Information.

No compliance issues identified. All required statements are present, specific, and internally consistent.