About the NRG
The Neuroinformatics Research Group builds tools and infrastructure to capture, curate, analyze, and share scientific data for the Center for Brain Science and Massachusetts General Hospital.
Web Apps for the Harvard Center for Brain Science:
DPdash is a deep/digital phenotyping dashboard developed and maintained by the Harvard Neuroinformatics Research Group. DPdash visualizes variables parsed from processed files that have been saved to a local file system (in our case, PHOENIX). Visualizing these variables allow users to verify that data are being captured successfully and explore relationships across data streams.
XNAT, originally developed at WUSTL, provides secure ways for researchers to archive, validate, and share MRI data. We have also developed pylib.xnat, a python library that provides tools to integrate XNAT into python scripts.
CBSCentral is our main installation, featuring BOLDQC2, which automatically generates quality reports you can use to make sure your data is consistent and identify scanning problems at the source. See the FAQ for details.
Other project-specific XNATs oriented more towards special projects (nrgcentral) or collaboration centers (contecentral, simons XNAT).
Ncfcode, a securely-hosted gitlab instance open to all CBS affiliates. This allows for GitHub-style collaborative coding on projects within the NCF. Unlike GitHub, this is level-3-secure and meets subject data security requirements such as those imposed by HIPAA. The NRG also uses ncfcode for continuous integration of software testing and deployment.
Web Apps for the Martinos Center at MGH:
onlinescoring.neuroinfo.org, a legacy system similar to redcap.
http://www.mytimedtest.org, a flash-based online cognitive ability battery.
We maintain the following XNAT installations at MGH, as well as https tunnels for NCF access:
embarc.dipr.partners.org for the Establishing Moderators and Biosignatures of Antidepressant Response for Clinical Care project.
gspcentral.dipr.partners.org for the GSP Data release.
nexuscentral.dipr.partners.org for Hedden lab..
PHOENIXThe PHOENIX filesystem is a secure, centralized, data warehouse configured to automatically import data from various web applications and mobile devices including XNAT, Beiwe, REDCap, OnlineScoring, iCognition, and Dropbox. A curated library of PHOENIX-compatible pipelines have been implemented and are available for users to download and extend. Many of these pipelines are configured to run automatically on all incoming raw data and any derived files are saved to designated processed folders for general use.
High Performance Consultations:
We can review your data analysis pipelines to make sure your scripts are running as efficiently as possible. For example, we were recently able to rewrite a fMRI warping script to be 30x faster, using some simple paralellization.
Module Creation and Custom Sofware Installations:
If there is a piece of software you can't install yourself, or which you think will be useful to the broader community, we can install it system-wide and create a module file for easy use by everyone on the NCF.
ContainerizationWe can help make your software more portable by containerizing it. This can also make dependencies more legible. We use Singularity for full compatibility with the Slurm batch scheduler. Singularity is currently available to NRG admin on a testing basis, and will be available to all users this summer.
- Logging into the NCF as an external collaborator
- See Users' version of this guide, with more links only accessible from NCF and partners network.
Open Data Releases:
- Brain Genomics Superstruct Open Data Release