Mouse brain reference atlases & mapping your own data

How do I map my data to your reference atlas?

We have been hearing this question (or variants of it) for as long as the Allen Mouse Brain Atlas has been online. The short answer is, We do not have a tool for this - but we have provided some resources to help, and many other researchers (and even companies!) have developed their own solutions. Some of those are shared here, and we welcome your additions to this list - either as publications, references to software or comments from other scientists who have experience in this area.

Tools from the Allen Institute:
First, view the information in the Software Development Toolkit (AllenSDK) on reference spaces. You can view and download file types, schema and protocols from here. Note that there are many types of reference atlases available (but we do not have a rat atlas!). Each atlas has a suite of analysis tools that range from very basic and minimal (e.g., spinal cord) to rich and complex (like the adult mouse brain, with a common coordinate framework or CCF). These templates don’t provide a mapping function as such, but you will need them to build your own. Unfortunately we do not have a service that can map our atlas to your data, or vice versa.

Resources from the community:
Many researchers have created tools for their own use; each has strengths and limitations in its applications. We do not endorse or support any of these, but provide links below for visibility.

  • QuickNII is a stand-alone software tool by Puchades, et al., for semi-automated affine spatial registration of sectional image data to a 3D reference atlas coordinate framework and can interact with the Allen Mouse Brain Reference Atlas.

  • HistoloZee is a software tool for interactively mapping 2D and 3D molecular and anatomical histology using Allen Mouse Brain Reference Atlas anatomical annotation.

  • NeuroInfo is a commercial product that MBF has created for mapping the Allen Mouse Brain Reference Atlas anatomical annotations to coronal sections of mouse brain that have a fluorescent label. Eastwood et al. use this method to map a mouse brain reference space.

  • Brainmapr is an application that the Kharchenko lab created for mapping data to the Allen Developing Mouse Brain reference atlas ontology.

  • Researchers at Genentech created an in-house analysis pipeline for mapping anatomical annotation from the Allen Mouse Brain Atlas to their Nissl-stained sections.

  • AtlasFitter is a Matlab based tool developed by the Brody lab, for the purpose of mapping annotation to experimental histological section images.

  • The Koshy lab also published a paper describing the creation of a Matlab application for the purpose of quantifying signal and mapping Allen Brain Atlas annotations to experimental data.

  • TissueMetrics provides a fee-for-service annotation mapping application for their clients, which includes lofting Allen Mouse Brain Reference Atlas anatomy to customer data.

  • Use of the Allen Mouse Brain Reference Atlas anatomical annotations is broad. View some of these cross-usages through Bioportal’s repository of ontologies.

  • The Scalable Brain Atlas utility initially developed by Bakker et al. includes the Allen Brain Atlas for mouse, and has functionality for importing user data for visualization.

What other resources are out there? What has worked (or not worked) for you?
Experts, speak up! Or share other conversations on the topic below.

1 Like

A Matlab pipeline to align histology images to the CCF (and then reconstruct probe tracts, etc):

Work in progress - feel free to fork/modify or ask questions


Also wanted to add our small contribution to the list:

We call it Semi-Manual Alignment to Reference Templates (SMART), an addition to Daniel Furth’s R package ‘Wholebrain’, that is specifically designed for aligning light sheet florescence microscopy datasets to the ABA CCF using minimal computational hardware requirments. Currently also documented in BioRxiv:


The in vitro physiology folks at Allen use this tool to make virtual slices from the CCF, which helps us to figure out which brain regions we are looking at in our slices:


Thanks Amy.

Just to say that cellfinder is still at a very early stage, but I want to get it working with as many people’s data as possible. There are more details at the github page. If you have any problems or feature requests, do get in touch there or via email.


1 Like