Whole Mouse Brain Modelling Workflow
Ero, C., Gewaltig, M. O., Keller, D., & Markram, H. (2018). A Cell Atlas for the Mouse Brain. Front Neuroinform, 12, 84. doi:10.3389/fninf.2018.00084
Goldowitz, D. (2010). Allen Reference Atlas. A Digital Color Brain Atlas of the C57BL/6J Male Mouse - by H. W. Dong. Genes, Brain and Behavior, 9(1), 128-128. doi:10.1111/j.1601-183X.2009.00552.x
Kyriakatos, A., Sadashivaiah, V., Zhang, Y., Motta, A., Auffret, M., & Petersen, C. C. (2017). Voltage-sensitive dye imaging of mouse neocortex during a whisker detection task. Neurophotonics, 4(3), 031204. doi:10.1117/1.NPh.4.3.031204
Oh, S. W., Harris, J. A., Ng, L., Winslow, B., Cain, N., Mihalas, S., . . . Zeng, H. (2014). A mesoscale connectome of the mouse brain. Nature, 508(7495), 207-214. doi:10.1038/nature13186
Figure 1: Whole brain scaffold workflow.
The whole mouse brain workflow is divided in 5 steps. The first step builds a Reference Framework from the Allen Mouse Brain Reference Atlas (Dong and The Allen Institute for Brain Science 2010). The second step consists in aligning volumetric mouse brain data, mainly from the Allen Institute for Brain Science (AIBS) to the Reference Framework, to form a volumetric Atlas. A positioning algorithm is used in step 3 to produce the Mouse Brain Cell Atlas, which holds our current knowledge of cellular composition of the mouse brain. The neurons of the Cell Atlas are connected according to additional data in step 4. Finally, electrical parameters are assigned to the neurons and their connections, allowing simulation of the resulting point neuron network.
Mouse Brain Reference Framework
Figure 2: Illustration of the Mouse Brain Reference Framework and its non-overlapping brain regions, colored according to the Allen Mouse Brain Atlas. Main brain structures (Left) such as olfactory bulb, cortex, or cerebellum can be divided in finer sub-structures (Right).
The reference framework defines the 3 dimensional volume of the mouse brain as well as reference volumes around 800 regions (see Figure 2). It will be used in HBP to replace mouse brain data in a spatial context, in order to create a scaffold atlas of our knowledge on the mouse brain.
Mouse Brain Cell Atlas
Figure 3: Creation of the Mouse Cell Atlas. Annotated Nissl stained slices of the mouse brain, are aligned and combined with ISH images to describe cell densities in each region of the mouse brain. A positioning algorithm is then used to place cells in the whole brain volume.
Figure 4: Illustration of the densities and absolute number of cells, neurons and glia in different regions of the mouse brain. The rings of each disk represent a hierarchical level in the Reference Framework, with higher structures closer to the center. The radial disk portions correspond to regions of the brain. Disks on the left shows the Literature information available on mouse brain cell densities and count, with missing data in gray. The disks on the right, displays the cellular composition described by the Cell Atlas.
After cell positions are determined, we label each position with a cell type (Erö et al., 2018). The cell-type is determined by integrating genetic information from in situ hybridization (ISH) image slices with region specific data from detailed circuit models, developed in HBP.
The labelled cell positions can then be used to reconstruct circuits of the mouse brain at different levels of abstraction.
Figure 5: Creation of the connectivity atlas. Long range connectivity is reconstructed from rAAV experiment slice images while short range connectivity is deduced from detailed circuit reconstructions. The resulting connectivity matrix links each cell of the mouse brain, in a complex neural circuit.
In the next step, we use two-photon tomography images of rAAV labelled axonal projections Allen Mouse Connectivity Atlas, to obtain the mesoscale connectivity between neurons in different brain regions (Oh et al., 2014). The intra-region connectivity can be predict from the combination of the types and positions of the neurons.
The whole brain connectivity can help to link different circuits of the mouse brain together, allowing the simulation of bigger regions and in the context of the entire brain.
Electrical Parameters Databases
Figure 6: Electrical features of a neuron and its connections. A pyramidal cell neuron from the Isocortex and its connections to the rest of the brain are displayed in the brain volume (left). Our database allows us to see its point neuron parameters depending on its electrical type. For each of its connections we can also retrieve the synapse parameters according to their synapse type.
To link cells and their connections to electrical behavior, we are constructing a database capable of storing multiple point neuron and synapse models together with their parameters. Tools will be also providing to explore this database and extend it.
The database can be extended as new data are provided by the HBP or literature. We are capable of storing parameters for every point neuron and synapse models. This will help greatly the HBP scientist to share and access point neuron network properties of the mouse brain.
Point Neuron Level Simulation
Figure 7: Simulation of the whole mouse brain after stimulation of a whisker barrel cortex. A, B, C shows propagation of neural activity of the whole brain model in respectively axial, coronal, and sagittal slices. The spiking neurons are displayed in color, according to the sub-regions they belong to.
Using the previously defined models, we are able to simulate the mouse brain as a point neuron network which will help us to further reproduce biological experiments, such as steady-state or after stimuli.
Figure 8: Virtual Voltage sensitive dye fluorescence during whole brain activity. We approximate voltage sensitive dye injection and fluorescence to be able to compare to experimental results such as (Kyriakatos et al. 2017).
We are creating a validation pipeline which will test our model and its simulation results against experimental datasets, from the cellular to the whole brain level, in order to get an objective quantification of the error generated by our models.
Figure 9: The whole brain model in the context of the HBP Co-Designed Project-1 (CDP-1). The workflow proposed by CDP1 is an iterative loop between experiments and simulation, which allows on one side to refine and validate models with experimental data and, on the other side, to redesign experiments based on simulations.
We want to embed the whole brain model in the whole mouse body in order to reconstruct the complete sensory-motor loop, collaborating with SP10 and as part of Co-Designed Project-1. This will help us to integrate behavioral data into our model but also to understand the interactions between central and peripheral nervous systems.
What makes a whole brain model special?
We build the whole brain model from the “outside-in” using data that describes the brain or parts of it, independent of any regional assumptions. This has the advantage that our model is not biased towards a particular brain atlas. Moreover our model can in principal be registered to different brain atlases.
What are the specific questions we are addressing in the HBP?
The model contributes to the detailed modelling efforts of HBP (SP6), by providing the global context into which detailed models of neural circuits or brain regions can be embedded.
In the context HBP Neurorobotics research (SP10) the whole brain model is connected to a virtual mouse body that is embedded in a dynamic environment. In the context of HBP’s cross disciplinary research (CDP-1) this embodied brain model is used to model in vivo experiments on motor rehabilitation after stroke. Ultimately, we aim at simulating behavioural experiments such as whisking and grasping. Future refinement of the whole-brain model therefore focuses on brain regions that are involved in sensory-motor processing.
What is our specific take?
In our work the workflow represents the ability to build the whole brain model in a routine basis.
It consists of a formalized cycle of modelling, validation and refinement. In each iteration of the cycle, a refined version of the model is produced, based on the information obtained during the validation results. Our workflow presents a significant improvement to current approaches, because it transforms the process of modelling from an ad hoc activity to a systematic operation.
First version of the Connectivity Atlas (by 2019)
First version of the Neuron model parameter database (by 2019)
First version of the Synapse model parameter database (by 2020)
Who is Involved?
Csaba Erő, Blue Brain Project, École Polytechnique Fédérale de Lausanne
Daniel Keller, Blue Brain Project, École Polytechnique Fédérale de Lausanne
Dimitri Rodarie, Blue Brain Project, École Polytechnique Fédérale de Lausanne
Marc-Oliver Gewaltig, Blue Brain Project, École Polytechnique Fédérale de Lausanne
Benefits to the Community
If you would like to use some of the results of this work (e.g. extract cell densities and types in specific regions of the brain), or get involved and contribute to our community effort, please contact any of us to discuss our common interests.