This dataset corresponds to a datamart produced by the WP2 team of the Landmark H2020 project. The database was developed by using a decision tree based script which determines the presence or absence of selected WRB diagnostic units (horizons, properties and materials) based on the harmonized soil profile dataset. The python-based code was developed based on the criteria defined by the World Reference Base for Soil Resources 2014 for the selected diagnostic units, by considering the difference in the information content of the input soil profile databases. Besides the presence/absence information, the code returns a percentage of reliability which provides an estimation on the reliability of the prediction of a certain diagnostic unit. The attributes are presented in the "dh_dictionary" file.