Main Menu (Mobile)- Block

Main Menu - Block

Positions

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

Recruiting software consultants

We are seeking in-person or remote software consultants to maintain and further develop our imaging tools, Cellpose and Suite2p. Cellpose and Suite2p are widely used in the biology community: Cellpose has ~8,000 downloads per month and has been cited over 300 times, with over 100 citations in 2022 alone; Suite2p has ~3,000 downloads per month and has been cited over 300 times. Therefore, work on these packages would reach hundreds of users across the world.

This is an exciting opportunity to be a part of a cohort of engineers dedicated to open source software development at Janelia. See more details here. Janelia also has competitive salaries and benefits packages. Please contact Carsen Stringer (stringerc@hhmi.org) with your CV and your github profile if interested. Also if you're interested in spike sorting and Kilosort, check out this job.

Job Description

The job has three main components: (1) the maintenance of Suite2p and Cellpose, (2) its integration in processing pipelines for Janelia labs, (3) the creation and documentation of new features developed in collaboration with labs at Janelia that will also be made available for the entire community.

For examples of new features to develop (3), here are features that have been requested by and/or are in progress in collaboration with Janelia labs:

  • 3D cytoplasm segmentation in large volumes from in situ RNA sequencing experiments
  • Simultaneous segmentation of nuclei and cytoplasm for improved segmentation for in situ experiments
  • Multi-day registration and cell detection from calcium imaging experiments for studying plasticity
  • Online cell detection for optogenetic photostimulation experiments
  • 3D registration and anatomical cell detection in the fly
  • Demixing functional activity from dense, overlapping cells in hippocampus and zebrafish
  • Anatomically-assisted functional segmentation of neural activity
  • Bleed-through correction for multi-channel imaging

Desired skillsets

  • Several years of programming experience in python and some experience with pytorch or tensorflow required, experience implementing automated testing preferred
  • Experience working with large-scale datasets and/or implementing machine-learning pipelines required
  • Experience developing new machine learning algorithms preferred

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

Suite2p is a fast, accurate and complete pipeline written in Python that registers raw movies, detects active cells, extracts their calcium traces and infers their spike times. Suite2p runs on standard workstations, operates faster than real time, and recovers ~2 times more cells than the previous state-of-the-art methods. Its low computational load allows routine detection of ~25,000 cells simultaneously from recordings taken with standard two-photon resonant-scanning microscopes. In addition to its ability to detect cell somas, the detection algorithm can detect axonal segments, boutons, dendrites, and spines. Suite2p has an extensive graphical user interface (GUI) which allows the user to explore their data, and is currently the only fully-functional pythonic GUI for calcium imaging data. Software developers have integrated Suite2p into their packages, such as those for multi-day cell alignment and photostimulation experiments.

Cellpose is a generalist, deep learning-based segmentation algorithm written in Python, which can precisely segment cells from a wide range of image types and does not require model retraining or parameter adjustments. Cellpose can be applied to 2D and 3D imaging data without requiring 3D-labelled data. To support community contributions to the training data, we developed GUI software for manual labeling and for curation of the automated results. We have retrained the model on community-contributed data to ensure the continual improvement of Cellpose. Software developers have integrated Cellpose into their own image processing software, such as CellProfiler, ImagePy, ImJoy, and aPeer. We also developed a Napari plugin for Cellpose (cellpose-napari).