Research tools and software

Free cloud platform for secure neuroscience data analysis.

Matlab toolbox to generate power-law colored noise signals of any dimensions.

Mathematica package for deterministic and stochastic simulation of Chemical Reaction Networks.

Developed with C++ and FLTK on a Mac to control a BVC-700 amplifier through ITC-18.

Simultaneously control a fast, low noise, and quantitative CCD camera through an EDT DV C-Link frame grabber board, and an amplifier through a HEKA ITC-18 data acquisition interface. The system is good for voltage imaging, calcium imaging, whole cell recording, and field potential recroding. Ephic is equipped with intuitive and versatile tools to control the experiment, and contains several powerful tools to analyze both imaging and electrical data.

Evaluate video clips of elevated plus maze test (EPM), which is a test to evaluate the anxiety level of rodents. EPM Analyzer counts the number of closed arm entries, the number of open arm entries, and time in the open arms. It also records the coordinates of the body center and the resident arm at each time point.

Quantify eyelid position during eyelid conditioning experiments.

Analyze videos of modified Forced Swim Test (FST) of animals.

Matlab toolbox to compute the statistics, pdf, cdf, inverse cdf and random numbers of the generalized chi-square distribution.

Perform digital image processing tasks such as enlargement (super-resolution), denoising, and realistic camera noise synthesis, using the statistical properties of natural images measured from large sets.

Matlab toolbox to integrate normal (Gaussian) distributions in any dimensions with any parameters within any domain, compute pdf/cdf/inverse cdf of any function of a normal vector, and compute quantities concerning classification performance among two or more multinormals, such as error matrix and discriminability d'.

Use generalized linear models to fit and analyze trial-based spike responses of multiple neurons with complex response properties.

Protocols published by the Harris lab on tissue preparation, microtomy, electron microscopy, 3D reconstruction and lab safety.

Analyze videos of Open Field Test (OFT) of animals.

We construct latent variable models of neural activity while modeling the relation between the latent and task variables. The resulting models are more constrained, and have better interpretability and identifiability. pi-VAE fits the data better, and also provides novel insights into the structure of neural codes.

A hardware architecture and Matlab software toolbox for neurophysiology that requires complex visual stimuli and online behavioral control.

Trace and annotate slice images, and combine them to create a 3D reconstruction. Reconstruct has been used to make 3D images of brain cells, blood vessels, the brain of an ant, human fetus bone, a sea lily, a fish fossil, metal alloys, tree bark, and spider genitalia.

Matlab code that implements the retina V1 model, which predicts the contrast detection threshold of a target in a background.

Publications related to space-variant imaging (due to foveation, or diseases such as glaucoma), information on foveated imaging, space-variant imaging software, and some examples of space-variant images and video.

Online, Mathematica and Python interfaces for answering questions about stable configurations of a Thermodynamic Binding Network.

SwiftCMA is a pure-Swift implementation of Covariance Matrix Adaptation Evolutionary Strategy (CMA-ES). CMA-ES is a population-based optimization technique that can optimize non-convex, non-smooth, non-differentiable functions. SwiftCMA is written in pure Swift, and makes proper use of functional programming and Swift's type system.

SwiftGenetics is a genetic algorithm library, written in Swift, that has been engineered from-scratch to be highly extensible and composable, by abstracting away different pieces of functionality, while providing concrete implementations of certain use cases, such as tree-based genomes.

Matlab function to trace sequential pixel coordinates of all contours in a binary image.

Tools to quantify the conceptual similarity of commonly known people, places, and things, using Wikipedia. Text from Wikipedia articles is extracted and embedded as vectors using the Universal Sentence Encoder, allowing the conceptual similarity of items to be quantified. Model estimates of conceptual similarity have been shown to predict human similarity ratings, and can be used to identify neural representations of conceptual knowledge.

Developed to test ITC-18, a computer interface to control other instruments, such as amplifiers, picospritzers, shutters, CCD cameras, etc. Programming support for ITC-18 is provided for IGOR and C/C++. One of the best ways to test the characteristics of a computer interface like ITC-18 is to simulate one waveform generator and one oscilloscope. WGO is developed with C++ and FLTK and can run on either a Mac or a PC.

The Zero-Inflated Gamma encoding and decoding model for post-deconvolved calcium imaging traces.