Main Menu (Mobile)- Block

Main Menu - Block

Positions

janelia7_blocks-janelia7_secondary_menu | block
janelia7_blocks-janelia7_fake_breadcrumb | block
Pachitariu Lab / Positions
node_title | node_title
Positions
node_body | node_body

Postdoctoral associates

The Pachitariu Lab at HHMI Janelia is looking for postdoctoral researchers to lead projects that investigate the neural basis of cognition in rodents. The lab uses large-scale neural recordings combined with modern computational and theoretical approaches. Learn more about our work here.

We are looking for researchers with a PhD in:

  • neuroscience / related field, particularly at the systems neuroscience level
  • computational field, with some experimental background in systems neuroscience

Some examples of projects ongoing in the lab include:

  • understanding how rodents perform complex visual tasks such as texture recognition and visual memory tasks
  • investigating the neural basis of few-shot learning and metalearning
  • interrogating the role of attention and behavioral representations in decision-making

The Janelia campus, including our lab, is committed to an inclusive environment for under-represented minorities and women. We strongly encourage applications from all under-represented groups. If interested, please send a CV to pachitarium@hhmi.org





Software consultants for Kilosort

We are seeking in-person or remote software consultants to maintain and further develop Kilosort, a set of spike sorting methods for electrophysiology. The ideal candidate for this position has several years of programming experience in python and also some experience working with electrophysiology data.

If you are interested in working on our imaging software (Suite2p, Cellpose), please see these positions instead. You can also apply to our scientific computing group (here), which has positions specifically for working on Kilosort and other open source project at Janelia (here). 

Job overview

This job entails a variety of engineering and research tasks: 

  1. Maintenance of the existing codebase with user support
  2. Prepare the full release of the Python version with some workshops to help the community transition
  3. Develop custom pipelines, in close collaboration with Janelia labs + release them publicly. The modular nature of Kilosort allows for many possible extensions.
  4. Help with analyses for the Kilosort paper. 

Diversity, equity and inclusion are important values at Janelia, and candidates should be dedicated to ensuring kindness and inclusion in their interactions with the open source software community and other employees at Janelia.

Software overview

Neuropixels electrodes, developed in a global collaboration between Janelia, IMEC and several other research groups, have single-handedly transformed the field of electrophysiology. Almost overnight, our community went from mostly using single-channel electrodes to mostly using 384-channel Neuropixel electrodes. Instead of painstakingly recording one neuron at a time like before, researchers can now record hundreds or thousands of neurons simultaneously. We are already witnessing the acceleration in research provided by Neuropixels, with Janelia at the forefront of its science as well.

Kilosort [paper1][paper2][matlab][python] is the main software used to analyze data from Neuropixels probes. Its main task is to take hundreds of GB of electrophysiology data and output a set of neurons with their spike times. The pipeline includes several distinct algorithmic steps, such as data registration, clustering, spike extraction and quality control. Read more about it in the original bioRxiv/Neurips paper and learn about the latest updates in the Neuropixels 2 paper. For a list of papers using Kilosort see this.

The hardware design of the Neuropixels probes has been strongly influenced by spike sorting results from Kilosort. For example, Neuropixels 2.0 has vertically-aligned columns of channels rather than the checkerboard design of 1.0, because we realized that vertical columns are much better for spike sorting. In turn, Kilosort 2.5 was optimized to perform well on such geometries. There is thus already a strong synergy between the software and the hardware sides of this highly successful Janelia effort. 

We estimate that Kilosort is used by more than 500 labs. This includes most of the labs using Neuropixels probes, as well as labs using other types of probes or even tetrodes. We don’t have a way to track downloads or unique installations, but the repo gets 500-1000 unique visitors / 2 weeks. This job will thus allow you to interact with a substantial portion of the neuroscience community.