DOME Registry Archived Dataset

Abstract This dataset is a comprehensive, time-stamped snapshot of the public DOME Registry database (registry.dome-ml.org). The DOME Registry stores annotations for supervised machine learning methods in biology, adhering to the community-wide recommendations for Data, Optimization, Model, and Evaluation (DOME).

Methodology This snapshot was generated using the "DOME Registry Downloader" tool. It queries the official DOME API (https://registry.dome-ml.org/api/review) to retrieve all public records, handling pagination and aggregation into a single structured file.

Content The dataset includes two files:

DOME_Registry_Download_[YYYYMMDD].json: The complete list of registry records in JSON format. Each record contains metadata regarding the ML study's data handling, optimization parameters, model architecture, and evaluation metrics.

DOME_Registry_Metadata_[YYYYMMDD].json: Technical metadata regarding the download process (timestamp, total record count, and file size).

Purpose This entry serves as a permanent, archival record of the registry state on the specified date. It is intended for researchers conducting meta-analyses on ML reporting standards or those requiring an offline version of the registry for reproducibility.