openMINDS stands for open Metadata Initiative for Neuroscience Data Structures. Within this initiative various metadata models are developed and integrated. These metadata models can be used as architectural building blocks in graph databases in order to register and describe heterogeneous neuroscience data structures.

Within openMINDS, a metadata schema defines properties needed to describe a certain data context (often reflected in the type or name of the schema) and the rules governing the value entries for these properties (e.g., the expected value type or the expected number of given values). A metadata model is established when multiple metadata schemas are linked. The rules governing the linkage possibility and type is defined within the schemas of a respective metadata model.

Each openMINDS metadata model covers a certain detail level or aspect around the description of neuroscience data structures. For example, the openMINDS_core metadata model is comprised of schemas that can be used to describe the general origin, location and content of neuroscience research products, while the openMINDS_SANDS metadata model contains schemas describing the anatomical locations and relation of neuroscience data. In order to facilitate the maintenance of the existing and the development of new metadata models, it was decided to keep them in different GitHubs governed by respective expert groups. Nonetheless, all openMINDS metadata models are interconnected and accessible as git submodules in a central openMINDS GitHub repository.

Generally speaking strings, integers, floats or booleans are derived data types. A string is defined as a sequence of characters between quotes (e.g., "Is this a string? YES!" or 'thisIsAlsoAString'). For a string, openMINDS accepts Unicode characters. An integer is a whole number, positive or negative, without decimals, of unlimited length (e.g., 5, -5 or 1238921234). A float represents a real number, written with a decimal point dividing the integer and fractional part, both of unlimited length (e.g., 5.15, -5.15 or 1238921234.1345). A boolean represents a logical proposition by means of the binary digits 0 (false) and 1 (true), especially in computing and electronics. How a boolean is written depends highly on the format or computational language.

A list is a data structure that is a mutable ordered sequence of values (also called items). The values of a list are typically defined between square brackets (e.g., [value1, value2, value3]). Note that the values within a list do not have to have the same data type. In contrast, an array is a data structure that is a mutable unordered sequence of values of the same data type. What data types are accepted for values in a list or in an array is highly depending on the format or computational language.

A key-value pair (sometimes also called name-value pair, attribute-value pair, property-value pair, or field-value pair) is a basic data representation and standard language feature in computing languages, systems and applications. In most cases this concept is used to build an associative array (also called dictionary), meaning an unordered list of unique keys with associated values typically defined within curly brackets (e.g., {key1: value1, key3: value3, key2:value2}). What data types are accepted for keys and values highly depends on the format or computational language. Note that a value could also be a data structure, such as a list, an array or an associative array.

JSON is short for JavaScript Object Notation, a lightweight data-interchange format which is built on associative arrays with key-value pairs and lists. Each JSON document/file begins as associative array. The keys are separated from the values via a colon and key-value pairs are separated by a comma. While a key always has to be a string in double quotes, a value can be a string in double quotes, an integer, a float, a boolean (written as true or false), null, a list or an associative array. Nesting of these structures is unlimited. For more information please go to the official webpage: https://www.json.org/. Several serialisation formats have been built on the JSON specification, such as JSON-LD (https://json-ld.org/). In addition, several schema languages have been developed to annotate and validate JSON documents, such as JSON-Schema (https://json-schema.org/) and SHACL (https://www.w3.org/TR/shacl/).