The engineers and data scientists of EDSI have contributed their expertise to a wide range of software; learn more about their projects by exploring the links and resources below.

Machine learning models for multi-molecular design

Post-doc: Chowdhury Ashraf , Mentors: David Beck and Jim Pfaendtner. A package for creating rich, equivariant, structural and chemical features from protein structure data. The ionic liquids design work software is available here and the membrane separations design software is available here.

FalseColor-Python project

Student: Robert Serafin, Mentor: Jonathan Liu. FalseColor Python is a rapid digital staining module designed for translating two-channel fluorescence images (i.e. a nuclear and cytoplasmic stain) to the traditional H&E histopathology color space. Website, technology and technique are available here.

Frameworks for ML of scanning probe images

Wesley K. Tatum, Diego Torrejon, Patrick O’Neil, Jonathan W. Onorato, Anton B. Resing, Sarah Holliday, Lucas Q. Flagg, David S. Ginger and Christine K. Luscombe. Machine learning on thin-film semiconducting polymer datasets. The m2py repo is available here and the py-conjugated repo is available here.

General machine learning software

Ariel Rokem and Kendrick Kay. Reparameterize ridge regression in terms of the ratio γ between the L2-norms of the regularized and unregularized coefficients. Open-sourced software implementations in Python and MATLAB are available here, and the software is available here.

Data-driven dynamics and control

Student: Henning Lange, Mentors: Steve Brunton and Nathan Kutz. Fourier and Koopman constitute spectral algorithms for learning linear and non-linear oscillators from data respectively. GitHub info available here.

Fluorescent image-based analysis

Joseph A., Liao R., Zhang M., Helmbrecht H., McKenna M., Filteau J., Nance E.

  • ifmodels – Use image registration between brain images of entire slices obtained via confocal microscopy and a universal atlas to create 3D immunofluorescent models of rat brains. GitHub info here.  
  • ifthresholds – A spinoff from ifmodels that uses thresholding, segmentation, and neuromorphology features to automatically score a variety of thresholds and machine learning to provide a ranked list of those thresholds for fluorescent cell images. GitHub info here.
  • fiberf – A basic repository to support our FIBER (Frameworks for neuroImage Based Experimental Routines) method in experimental and data science collaborations on analyzing neuromorphology.  Specifically used to study cells from ferret slices. GitHub info here.
  • textile – A GitHub Repository to contain lessons and modules for our TEXTILE (Tutorials for EEperimentalist Interactive LEarning): Data Science in Biological Sciences educational course for lab trainees. Includes pre-module activities, modules, and recall activities to learn both data science and image processing from a biological researcher perspective. GitHub info here.
  • cellmorphflows – A repository to hold our single experiment Jupyter notebook workflows for running the VAMPIRE process and analysis. GitHub info here.