Computational Materials Science: Data in Brief

Two complementary initiatives: Open Data and Data in Brief

Research data in materials science is often inaccessible or buried in the supplementary content of an article. In line with Elsevier’s activity to promote open research data, we encourage authors to share peer-reviewed, curated, formatted, indexed, citable raw research data, made publicly available to all upon publication. Computational Materials Science is now the first journal in the materials science portfolio to offer two options to facilitate the sharing of research data.

First, Computational Materials Science is one of the few journals, alongside CALPHAD, to make use of Elsevier’s new Open Data service which allow authors to place datasets directly which are hosted alongside their articles on Sciencedirect.

Prof. Susan Sinnott, Editor-in-Chief of Computational Materials Science, said: “I am pleased that Computational Materials Science is now participating in Elsevier’s new pilot: Open Data as this service provides authors with the option to easily upload their raw research data as a supplementary file free of charge.”

Second, Computational Materials Science now directly links to Data in Brief, which is a new Open Access journal, where researchers can publish a short article (termed Data in Brief) describing one or more data sets that underpin the study reported in their linked article. The authors can make the data publicly available either through a repository of their choice or directly alongside the data article, and the Data in Brief provides valuable context and metadata needed to facilitate reuse of the data.

Linking to Data in Brief

The first Computational Materials Science research article exploiting this new link with Data in Brief was authored by Dr Rivero and Prof. Salvador Barraza-Lopez, from the University of Arkansas, and his co-workers and is entitled Systematic pseudopotentials from reference eigenvalue sets for DFT calculations and can be freely accessed for the next few months, or downloaded from Materials Today.

The corresponding Data in Brief can be along here: Systematic pseudopotentials from reference eigenvalue sets for DFT calculations: pseudopotential files along with the associated data, which consists of new pseudopotentials to be used for density-functional theory calculations, which were uploaded by the author directly with the data article.

As pointed out by Prof. Barraza-Lopez: "Pseudopotential files are a delicate input for density-functional theory calculations. The files were generated over a few months, and tested in more than one atomistic configuration. But it should be clear that the ultimate test rests with the users of the SIESTA code: having these input pseudopotential files freely accessible, they can be used and to be further validated by the community at large."

Both the data and the article are fully Open Access and can be re-used in compliance with a CC-BY license.

Prof. Barrada-Lopez indicated that “with Data in Brief, many developments in research can become more useful when data sources are shared. We are excited and grateful for the opportunity to have our data accessible at no cost to the community."

Data in brief

Data in Brief allows authors to provide a precise description of the data as well as share the metadata necessary to facilitate re-use of the data by filling in a simple template. Authors can place the datasets alongside the data article as supplementary material, in a repository of their choice or make use of the Data in Brief Dataverse hosted by Harvard University. To make it easy for authors, the can be uploaded directly with their submission to Computational Materials Science via the editorial system, which guarantees a greatly reduced Open Access fee.

Open Data on ScienceDirect

How does open data work?

1. At the submission phase, in EES, authors upload their raw research data as a supplementary material and classify the file as "raw research data," which is validated after submission.

2. After acceptance, the data file is made available open access to both subscribers and non-subscribers to view and download from ScienceDirect under a CC-BY license.

For Computational Materials Science, what types of data are included?

A broad range of data will be accepted. This includes optimized material structures, e.g. indicating lattice structures as well as microstructures, parameters and functions for models, software, and aggregated positions and reactions from the analysis of simulations.

For Computational Materials Science, what are the data and metadata standards?

Authors are encouraged to provide data in standard formats for the type of modeling used, for instance POSCAR for VASP, .in files for GULP, .m or .mat files for Matlab, data analysis files as well as software in the form of code or compiled files.

Computational Materials Science is the first to have both these features, more journals will soon follow as we believe in supporting the move towards Open Data and aim to do this in the best interest of our authors.