A Molecular Atlas for the Brain

© 2021 EPFL

© 2021 EPFL

Blue Brain open sources a simulation-ready database to accelerate molecular and systems biology.

Molecules are the fundamental functional blocks of all processes that occur within cells. All brain functions, including cognition and the senses, rely on complex molecular chain reactions occurring within neurons and these processes, are ultimately governed by many thousands of molecules interacting every microsecond. Diseases stem from abnormal molecular levels and drugs try to correct these interactions.

In a paper published in Frontiers in Molecular Neuroscience, Blue Brain researchers discuss how accurate molecular concentrations are essential to reliably study biochemical networks and, in the creation of predictive models for molecular and systems biology. Both protein and metabolite concentrations used in current models are often restricted or irreproducible and are not measured under the same conditions. Inconsistencies in reported data fuel misconceptions of the molecular contents of cells and how they interact. Data from different sources also give rise to conflicts in nomenclature (devising or choosing of names) and the use of different units of measure, as well as discrepancies in experimental procedures, data processing and implementation in models.

Demonstrating the importance of a data-driven approach for modeling and simulation

To overcome these obstacles that have held molecular and systems biology back for decades, Blue Brain scientists performed an extensive meta-analysis to integrate protein and metabolite quantitative data from publicly available resources and created a simulation-ready database to support more standardized and comparable molecular and systems biology studies – the Brain Molecular Atlas.

“Our data integration provides a resource for the study of the brain, as well as guidance for future experiments and data science. Despite the global effort in the standardization of data, the differences in experimental procedures, nomenclature and even units distort the reuse of publicly available data,” explains PhD student Polina Shichkova, lead author of the study. “The meta-analysis study allowed us to bring together knowledge from various sources to understand which information is already available and what can we learn from it, and which strategic data needs to be obtained in future experiments. Our approach can be adapted to any biological tissue from any species and the paper highlights aspects of critical data assessment for more general use. Altogether, we demonstrate the importance of a data-driven approach for modeling and simulation with species, brain region and cell-type specific molecular concentrations,” she concludes.

A scaffold, which can constrain other aspects of brain composition

Daniel Keller, Head of the Molecular Systems group at Blue Brain, explains further the importance of linking the molecular composition to defining the cell type. “The brain is a tapestry of different cells, and each type of cell has a different complement of molecules. A standardized approach to defining cell types based on their molecular profiles should see more consistency in the properties and characteristics of neural tissue.”

Providing more complete molecular profiles of brain cells

“These findings have enabled us to expand the variety of molecules we consider at Blue Brain as we near completing a simulation of the entire mouse brain,” says Blue Brain Founder and Director, Prof. Henry Markram. “The list of ‘classic’ ions and proteins studied in neuroscience is quite limited to those directly involved in neuronal signaling related functions. There are many more biochemical pathways with corresponding proteins and metabolites therein. For the neuroscience community, access to a more complete molecular view of what is inside brain cells are particularly important for translational studies related to brain disorders, where a significant knowledge about the contribution of particular proteins to the brain malfunction is accumulating,” he explains.

The standardized Brain Molecular Atlas has been open-sourced as a resource for systems modeling with a key element being that it overcomes the obstacles of missing or inconsistent data to support molecular and systems biology. It can also provide the baseline for healthy concentrations, for concentrations that reflect disease and how these concentrations change during aging, thereby providing a valuable resource for data science and modeling brain metabolism as well as guidance for future experiments.


Citation – Shichkova P, Coggan JS, Markram H and Keller D (2021) A Standardized Brain Molecular Atlas: A Resource for Systems Modeling and Simulation. Front. Mol. Neurosci. 10 November 2021 | https://doi.org/10.3389/fnmol.2021.604559

Brain Molecular Atlas

The Brain Molecular Atlas is the first step to providing a resource for the detailed data-driven reconstruction and simulation of the molecular processes in the brain. As more data becomes available, it will be refined and expanded. The integrated data in the Brain Molecular Atlas is publicly accessible through the Blue Brain Cell Atlas (Erö et al. 2018) and Subcellular Protein Atlas of brain cells application. Both can be accessed here - https://portal.bluebrain.epfl.ch/resources/models/brain-molecular-atlas/

For more information, please contact Blue Brain Communications Manager, Kate Mullins


This study was supported by funding to the Blue Brain Project, a research center of the École polytechnique fédérale de Lausanne, from the Swiss government’s ETH Board of the Swiss Federal Institutes of Technology.