Scientist I – Modeling Structure-Function Relation in a Reconstructed Cortical Tissue

Employer
Allen Institute
Location
Seattle, Washington State
Salary
Competitive Salary
Posted
October 16 2020
Position Type
Full Time
Organization Type
Non-Profit

Scientist I – Modeling Structure-Function Relation in a Reconstructed Cortical Tissue

Our mission at the Allen Institute is to advance our understanding of how the brain works in health and disease. Using a team science approach, we strive to discover how the brain implements fundamental computations through the integration of technological innovation, cutting-edge experiments, modeling, and theory.

Understanding how the structure of biological neuronal networks leads to its observed activity, and how it relates to the implemented computations is one of the primary challenges in computational neuroscience. We have large datasets of in-vitro measured cell type properties (https://celltypes.brain-map.org/), statistical knowledge of their connections (https://portal.brain-map.org/explore/connectivity/synaptic-physiology) and large-scale recordings of their in-vivo activity (https://observatory.brain-map.org/visualcoding). These are complemented by a fantastic dataset of coupled structure-activity measurement at cell type level in one tissue (https://microns-explorer.org).

We are seeking a scientist to join the Modeling and Theory team of Stefan Mihalas help construct data driven models of the activity and computations in the mouse cortex. Our team focuses at extracting principles from large data, integrating them in simplified models and using the models to test theories of computation.

Even with all this data available, directly trying to simulate activity from structure is an under constrained problem. The scientist will work to compare existing algorithms to efficiently explore the space of models linking structure to activity and potentially develop new algorithms, if needed. The scientist will explore how additional constrains, including constraints imposed by computations, can be used to further narrow down the parameter space. To be able to explore the space of models efficiently, we intend to use on spiking neuron models, with a focus on generalized leaky integrate and fire models. The scientist will work as part of a team to use these methods to construct a set of models visually driven in vivo neuronal activity, and subsequently explore the computational properties of these models.

Essential Functions

  • Implement and compare algorithms for automatic parameter optimization in spiking neural network models
  • Help design and construct data driven models for visually driven in vivo neuronal activity
  • Help explore the computational properties of the models
  • Contribute scientific ideas based on the modeling results
  • Develop and maintain computational and associated software tools
  • Publish/present findings in peer-reviewed journals and at scientific conferences
  • Maintain clear and accurate communication with supervisor and other team members

Required Education and Experience

  • PhD in computational neuroscience, applied mathematics, computer science or a related field
  • 0-2 years of post-doctoral experience.
  • Strong computational/data analysis skills and good programming ability in Python
  • Strong written and verbal communication skills

Preferred Education and Experience

  • Strong knowledge of systems neuroscience
  • Experience using PyTorch or TensorFlow
  • Experience simulating spiking neuronal networks using NEST
  • Ability to work as part of a collaborative team
  • Track record of scientific excellence and independent thinking