The MAX Centre of Excellence (CoE) aims at supporting the needs of all the stakeholders involved in the field of materials modelling, simulation and design by providing new instruments and services in the form of data, codes, expertise and turnkey solutions to efficiently address the crucial challenges of novel materials development in the exascale computing era. This document provides a description of the strategies and solutions adopted within the MAX CoE to establish a high level materials’ informatics framework to curate, preserve and share all the data produced by the flagship codes.The core technology behind this objective is the AiiDA code, a python infrastructure designed to support different codes through a plugin interface, allow for an automated design and implementation of complex workflows and task tracking and able to store the full provenance of each object in a tailored database. AiiDA parses the input and output files and runs the calculations on high performance computing platforms, stores the data using uniform formats based on python dictionaries and preserve the full provenance in the form of a Directed Acyclic Graph (DAG).AiiDA also enables a social ecosystem where the simulation workflows and results can be openly shared, on one hand with the update of the AiiDA plugin and workflow systems and on the other, with the development of the AiiDA REpresentational StateTransfer (REST) Application Programming Interface (API) which also constitutes the backbone of the Materials Cloud portal and finally, with the implementation of various exporters and converters to the most commonly used data formats and ontologies. Longterm sharing and preservation is supported thanks to the Materials Cloud Archive open repository, guaranteeing storage for at least 10 years after publication. The Archive is integrated with the Materials Cloud Explore section, that guarantees a Findable, Accessible, Interoperable and Re-usable (FAIR)-compliant sharing of data produced by AiiDA, and with the Materials Cloud Discover section, for sharing highly curated data. Statement on Open Research Data: In MAX , we believe that sharing research data in a FAIR format is crucial to guarantee reproducibility, increase transparency and impact of research and accelerate discovery. For this reason, except when data is bound to confidentiality by legal, ethical or copyright reasons, MAX researchers will deposit data needed to reproduce a scientific paper published within the scope of the MAX CoE on the Materials Cloud Archive with open licenses, guaranteeing that everybody can find, access and reuse the data without restrictions.
Nicola Marzari (Fri,) studied this question.