Software availability and
installation
PatientMatcher is open-source and available on GitHub
(https://github.com/Clinical-Genomics/patientMatcher).
The software is distributed under the MIT license
(https://github.com/Clinical-Genomics/patientMatcher/blob/master/LICENSE)
and we encourage all interested parties to use and modify its code
according to their needs. The main GitHub repository is curated by
Clinical Genomics, but we look forward to establishing a collaborative
environment where other users could help improving the code, adding or
simply requesting new useful features.
The simplest way to run and test the server is to use the up-to-date
container image with a basic software installation that can be pulled
from Docker Hub
(https://hub.docker.com/repository/docker/clinicalgenomics/patientmatcher).
On the GitHub pages of the repository, we also provide instructions and
support files to test PatientMatcher with real data without needing to
install any software, except Docker. For this purpose, we compiled a
multi-container Docker Compose file that, when launched by a single
command from the terminal, provides a complete setup of the server,
including a running instance of MongoDB containing the 50 benchmarking
patients spanning 22 disorders described in Buske et al.(J. Buske et
al., 2015). This server represents a standalone MME instance, ready to
accept HTTP requests on localhost and port 9020. For development and
testing reasons we have also created a more sophisticated Docker Compose
setup, with an MME server connected to another two MME nodes (other
instances of PatientMatcher), both containing demo patient data. This
file is available under the /containers folder in the GitHub page of the
PatientMatcher software. Deploying the software in a production
environment using the official Docker image file could be achieved using
Kubernetes (https://kubernetes.io/) or via
Podman (https://podman.io/) system services. Another tested way to
deploy the software is installing it from the Python Package Index
(PyPI) using the Python installer Pip. In this case it is recommended to
operate in a virtual environment, such as Conda
(https://docs.conda.io/) after installing Python 3.6+. All these
options, together with other server settings, are extensively described
on the software GitHub pages.