What purpose is the DDMoRe library serving?

DDMoRe, Drug Disease Model Resources, (http://www.ddmore.eu/) is an European project founded by the IMI and EFPIA, which aims to improve efficiency in the process of Model-informed Drug Discovery and Development (MID3). One of the key products is the DDMoRe model repository, an open, publicly-available and free-to-use cloud-based model library for drug and disease models currently under development and which will be further supported by the DDMoRe foundation. Among the current bottlenecks in applying efficiently MID3 is the complexity of the structural and statistical sub-models and its coding, as well as dataset arrangement. In addition, drug and disease models as they are currently available in the scientific literature do not permit direct comparisons across models, and are often difficult to re-implement without code listings, which are uncommon. Therefore an efficient storage of models covering the majority of the needs in pharmacometrics practice is an urgent need, and has been one of priorities of the DDMoRe activities. To date no publicly-available drug and disease model libraries exist except for recent private initiatives To fulfil part of the daily requirements in pharmacometrics an efficient repository needs to be far beyond a simple warehouse. Under this philosophy the DDMoRe model repository provides a platform to: (i) store and share models in a secure and version control manner, (ii) friendly display key model features (e.g. structure, parameters, initial conditions), (iii) annotate models based on ontology standards and metadata, and (iv) connect to external databases opening opportunity for access from scientists of other life science domains. Those main characteristics together with a fully searchable engine can certainly increase the quality of communication and dissemination of results and advances among the pharmacometrics community. The repository has been prepared to support models encoded in MDL and PharmML, the two standard languages developed by the DDMoRe consortium to enable interoperability across different modelling & simulation softwares and to facilitate communication of M&S models across different disciplines. Nevertheless models provided in original target code are also accepted. In order to publish a model in the repository a set of minimal requirements, automatically checked when submitting a model, needs to be met. These requirements include the availability of an executable model code, an example of real or dummy dataset guiding data organization, a command text file illustrating how to execute the model, model output results and a set of annotations describing the modelling exercise and its context of use. The goal is to warrant the comprehension and re-usability of models. Currently, the repository contains models related to pharmacometrics in areas such as oncology, diabetes, central nervous system or infectious diseases. Models are supported by the corresponding publication in peer review journals or conference abstracts, but it is also open to models from other fields such as system pharmacology/biology. In summary, the DDMoRe model repository provides the ground for a successful library which usefulness will ultimately depend on the community acceptance and adoption. It should be noted that currently the quality of the information available depends solely on the submitter. A certification process is currently under development to initially qualify published models from a technical point of view to ensure model correctness and re-usability; future extensions will consider establishment of a scientific validation.