NiMARE is a Python package for performing meta-analyses, and derivative analyses using meta-analytic data, of the neuroimaging literature. While meta-analytic packages exist which implement one or two algorithms each, NiMARE provides a standard syntax for performing a wide range of analyses and for interacting with databases of coordinates and images from fMRI studies (e.g., brainspell, Neurosynth, and NeuroVault).
NiMARE joins a growing Python ecosystem for neuroimaging research, which includes such tools as Nipype, Nistats, and Nilearn. As with these other tools, NiMARE is open source, collaboratively developed, and built with ease of use in mind.
This page outlines NiMARE’s purpose and its role in a proposed meta-analytic ecosystem.
A Proposed Meta-Analytic Ecosystem¶
NiMARE aims to fill a gap in a burgeoning meta-analytic ecosystem. The goal of NiMARE is to collect a wide range of meta-analytic tools in one Python library. Currently, those methods are spread out across a range of programming languages and user interfaces, or are never even translated from the original papers into useable tools. NiMARE operates on NIMADS-format datasets, which users will be able to compile by searching the NeuroStuff database with the pyNIMADS library. A number of other services in the ecosystem will then use NiMARE functions to perform meta-analyses, including Neurosynth 2.0, NeuroVault, and metaCurious.
This page outlines a tentative plan for a system of services for neuroimaging meta-analysis. Several of the services detailed here do not currently exist or only partially support the functionality described below. This plan is likely to change over time.
Neurosynth currently stores a coordinated-based database of over 14,000 neuroimaging papers (automatically curated by ACE), provides a web interface for automated meta-analyses, functional decoding, and gene expression visualization, and provides a Python package implementing the above methods.
In order to improve modularization, the next iteration of Neurosynth will limit itself to a web interface for meta-analytic model specification and providing a centralized storage for large-scale meta-analyses, but not actually implementing the algorithms used to run those meta-analyses or to perform the other services provided on the website (e.g., functional decoding and topic modeling). The algorithms currently implemented in the Neurosynth Python package will be implemented (among many others) in NiMARE. Under the current plan, the database at the moment stored by Neurosynth will instead by stored in the NeuroStuff database, which will also store other coordinate- and image-based meta-analytic databases in NIMADS format.
brainspell is a clone of the Neurosynth database meant for crowdsourced manual annotation. It provides a website where users can correct mistakes made by ACE or can add labels from multiple cognitive ontologies (including the Cognitive Paradigm Ontology and the Cognitive Atlas) to experiments.
metaCurious is a new frontend (i.e., website) for brainspell, oriented toward meta-analysts. MetaCurious provides search and curation tools for researchers to build meta-analytic samples for analysis. Search criteria, reasons for exclusion, and other labels may be added by the researcher and fed back into the underlying database, resulting in goal-oriented manual annotation. MetaCurious generates GitHub repositories for meta-analytic samples, which will also be NiMARE-compatible in the future.
NIMADS is a new standard for organizing and representing meta-analytic neuroimaging data. NIMADS will be used by NeuroStuff, pyNIMADS, metaCurious, and NiMARE.
NeuroStuff (tentatively named) will act as a centralized repository for coordinates and maps from neuroimaging studies, stored in NIMADS format. Users will be able to query and add to the repository using its API and the pyNIMADS Python library.
pyNIMADS (also tentatively named) is a planned Python library that will act as a wrapper for the NeuroStuff API, allowing users to query the database and to build NiMARE-compatible datasets for analysis.