Physical and Computational Chemistry

Introduction

Physical Chemistry encompasses a variety of sub-disciplines, covering vast methodology, which, in turn, produce heterogenous data. From spectroscopic measurement data, imaging, to simulation input files and in-house data analysis code, many physical chemists are experienced in handling digital data. They are often well-versed in developing software solutions to support their work. While some work with large data volumes on a regular basis, other methodologies result in small, text-based files. This discipline includes many data-literate members, the expertise of which may be harnessed to implement tools and solutions to manage their research group's data in a unified and streamlined manner.

Data Types

As mentioned, the data produced in physical chemistry and its diverse sub-disciples are varied. One research group may work intensively with imaging data such as superresolution microscopy, while the other may work on method development, and again others may analyze spectroscopic data or conduct numeric simulations—or even any combination of these.

ELNs and Other Tools

For effective data management, software tools should be selected in a uniform manner within a project or research group with the aim to organize and streamline workflows. This involves establishing clear usage guidelines, including metadata templates drawn from minimum information standards for a given method, where available. These should be outlined in a data management plan (DMP) for each project. Many universities supply tools and templates for DMPs (see the respective article for more information).

An electronic lab notebooks (ELNs) helps in the day-to-day planning and structured documentation of experiments, while some also assist in data workflow management. For disciplines with diverse research, ELNs must be flexible and customizable. Certain universities may have a central option, while each research group may chose what best fits their needs and resources if they are able to host or procure their own solution. The ELN-Finder lists many options and the article on choosing an ELN provides further assistance:

Loading...

In addition to ELNs, tools such as local repository and research dat management tools can assist in making data publication ready.

For those writing scripts and developing research software solutions, Git is a highly recommended versioning tool. May universities also have their own instances of GitLab to assist in managing software projects.

Specifically for research data, DataLad, which is built on top of git, can greatly assist in tracking the metadata while processing and analyzing data. While it works for steps carried out with GUI applications, its true power comes in handy for those using script-based analysis and processing steps.

As many physical chemists may establish their own workflows or develop their own tools to acquire, process, and analyze their data, it is highly recommended to adhere to community-specific standards for file formats and metadata, where available. A bare mininum is to establish documentation and format standards within a research group, ensuring an efficient knowledge transfer from one generation of researchers to the next. An ELN can greatly assist in providing templates for documentation, while a DMP should be used to record the format standards. Employing automated workflows, e.g., from device to ELN and data storage systems, can greatly reduce manual steps in day-to-day work and can automatically ensure data and documentation are complete and formatted correctly. Some tools provide out-of-the-box solutions, such as device integration, while others provide options such as REST APIs to build custom methods.

Publishing Data

Publishing research data, especially that underlying a published article, is an important aspect that allows others in the research community to replicate and build upon a researchers work. Research data repositories serve as platforms for data publication and can greatly assist in FAIR data publication. Such repositories range from subject-specific to general and institutional. For many in physical chemistry and its varied data, general repositories such as RADAR4Chem presents an option for publishing data for which now (sub-)discipline-specific repository has been established. ioChem-BD serves as a computational chemistry repository and includes a conversion service for many common data types to ensure interoperability. The Image Data Resource (IDR) makes biological imaging data available to the community, while self-hosted Omero repositoy can assist those working with other types of imaging data. For more information on choices, head here.

For work that includes developing in-house research software solutions: software is data and an integral part of research and should be published as such. While only GitHub currently offers an automatic workflow for publishing software releases to zenodo, there are methods to assign the software a DOI, therefore making it citable.

Challenges

Common challenges in physical chemistry and FAIR data often go hand-in-hand with the large variety of sub-disciplines, methodology, and thus diverse data types. Many working in physical chemistry labs may have established their own personal workflows. Working within a group streamline and unify common steps and to establish reusable templates for metadata, be it in ELNs or in the local file system, can provide structured information not just for fellow researchers, but also for those working on FAIR data infrastructure, such as ELNs and research data repositories.

Especially in imaging, large data volume can strain the local storage resources. Central storage solution can provide assistance and should be used in combination with best research data management practices to ensure the data's re-usability and avoid unorganized and inefficient use of large storage systems.