ankursinha.in/blog

neuroscience/fedora/musings

Current Projects

I am a post-doctoral Research Fellow at the Silver Lab at University College London working on computational modelling projects. My work focuses on building tools and standards for biophysically detailed computational modelling in neuroscience and am actively involved in several open source communities.

Research & development

Investigation of cortico-cerebellar loops

https://github.com/sanjayankur31/GolgiCellNetwork

This involves developing biophysically detailed models of components of the cortico-cerebellar loop to investigate their functions and mechanisms.

  • Status: active development
  • Tools: NeuroML, Python, NEURON, NetPyNE

Standardisation of Human L2/3 cortical model

https://github.com/sanjayankur31/Human-L2-3-Cortical-Microcircuit

  • Status: model conversion complete, network validation in progress
  • Tools/technologies: NeuroML, Python, NetPyNE

This involves standardising the published Human L2-3 cortical microcircuit model to NeuroML for community use:

  • ion channels
  • neuronal morphologies
  • synaptic mechanisms

NeuroML

2020 -

https://docs.neuroml.org, https://github.com/NeuroML

NeuroML is an international, collaborative initiative that maintains an ecosystem of tools for biophysically detailed/realistic neuronal modelling. The ecosystem consists of:

  • a simulator independent specification (standard)
  • core tools for reading/writing NeuroML (APIs), and running simulations using different simulation engines
  • an extended ecosystem of software tools that are compliant with the standard

I have served as an elected editor (core maintainer) of the NeuroML standard and software stack since 2022. This includes:

  • maintaining the standard
  • maintaining the core APIs and software tools
  • maintaining/adding support for simulation engines, collaboratively with the help of the maintainers of these engines
  • undertake outreach, dissemination, and training activities

LLM/AI assistance

We are currently exploring methods by which we can leverage LLMs/AI tools to accelerate the building, fitting, simulation, and analysis of NeuroML models, especially to make modelling more accessible to the non-technical/non-computational neuroscience research community.

  • Status: prototyping
  • Tools/technologies: Python, Langchain, RAG systems

Open Source Brain

2020 -

http://www.opensourcebrain.org/, https://github.com/OpenSourceBrain/OSBv2

  • Status: stable
  • Tools/technologies: Python, Javascript, React/Redux, OpenAPI, Kubernetes, Docker, Google Cloud Platform

Open Source Brain is a cloud based integrated research platform for collaborative data analysis and computational modelling.

I am a core team member that maintains the platform and manages its development.


Community & leadership

OCNS

2019 - 2021, 2023 -

https://www.cnsorg.org

I am an elected member of the Board of Directors for Organisation for Computational Neuroscience (OCNS). I was previously the "webmaster", in-charge of the website and infrastructure.

I am now in-charge of the OCNS newsletter.

I also co-chair the INCF/OCNS Software working group. We focus on all things related to the software used in Neuroscience---development, training and dissemination.


Fedora Linux

2008 -

https://fedoraproject.org

  • Tools/technologies: Python, C/C++, Rust, RPM, Koji, Bash, Discourse, IRC, Matrix

I maintain >300 packages for the Fedora Linux distribution. I am also a sponsor to the package maintainers team, and have "proven packager" access to all community packages.

I am a core member of the NeuroFedora team. Our goal is to provide a large set of neuroscience software to users in an easy, ready to use way---via the Fedora Linux distribution.

I am also a core member of the Fedora Join SIG that helps to onboard new community members.


Others

I also contribute to a number of other neuroscience and Open Source software projects that can be seen on my GitHub profile.