MT LIP GLM

This Repo provides MATLAB code for the analyses in Yates et al. (2017)

Getting started / installation

Installing the code

Create a local copy of mtlipglm by cloning or downloading a zip

	git clone https://github.com/jcbyts/mtlipglm

You will also need the classy branch of neuroGLM code

	git clone https://github.com/jcbyts/neuroGLM
	cd neuroGLM
	git checkout classy

Getting the data

The code will not run without local copies of the data. To quickly reproduce figures from the manuscript requires both the data and the fitted GLMs. Because all the data and fits takes up a fair amount of disk space, there are a number of options.

You can download the data through #1 or #2 and then refit yourself, or download #3 as well for a quick reproduction of figure 5.

If you downloaded the fits, move main_fits and lip_trunc_fits to the directory where neurons and stim are. The mtlipglm code will assume that they are all in the same base directory.

Before running

Note: this code has only been tested on MATLAB 2015b and newer.

setup paths

Open Matlab. Change directories to mtlipglm. You need to set up the specific paths for your machine.

edit addMTLIPpaths

Find the lines

% this is where your data live
dataPath = '';

and

% neuroGLM - must be downloaded from https://github.com/jcbyts/neuroGLM 
neuroGLMpath = '';

and enter the appropriate path locations for the base data directory and the neuroGLM repository.

run addMTLIPpaths and if setup correctly, then

ls(getpref('mtlipglm', 'dataPath'))

should show you the neurons, stim, main_fits, and lip_trunc_fits directories.

Running the code

This repository can reproduce figure 5. It has all the fitting code for fixed hyperparameters. It also has an example script for exploring the data and creating new GLMs.

Figure 5 b-f

If you have downloaded the fits, either run figure05 or open it and run cell by cell. This will reproduce figure 5, panels b-f from the manuscript.

If you have not downloaded the fits, you will need to run cmdFitLIP. This will probably take some time (30min+) to refit all the required models with a fixed hyperparameter. Once this is completed running, you can run figure05 to plot the results.

Figure 5 g,h

To reproduce the effects of elongating the acausal portion of the pre-saccadic covariates, run cmdFitLIPtrunc. If you downloaded the fits, this will re-analyze them and produce the required data for plotting results. Otherwise, if will refit each model. You can adjust the number of truncation steps to make this faster.

Example

To explore the data / try alternate GLM parameterizations,

edit cmdExample.m

Basic Overview

There are a few classes that govern the analyses in mtlipglm.

Try running

dataPath = getpref('mtlipglm', 'dataPath');
neurons = getNeurons('p20140304', dataPath)

You'll notice that neurons is an array of neuro objects. This is the most memory efficient way to access the neuron data. Each neuron file is an hdf5 file that includes many attributes about the unit under study (e.g., receptive field maps, waveform properties). Rather than load all of these files and atributes into memory, the neuro object can access each attribute when needed. To get the full struct for a neuron, just run

n = neurons(1).getStruct;

or load the file directly. However, as mentioned, the code is going to use the neuro class, both because of its lazy access properties, but also because it has a number of useful methods for getting more info about the neuron. The neuro class has a property stim that similarly accesses the stimulus properties when needed.

The rest of the code revolves around two classes: mtlipglm and glmspike

glmspike is an inherits neuroGLM and can be used the same way. It contains additional methods for setting up particular parameterizations of the basis functions, compiling the design matrix, and performing the fitting.

mtlipglm is the workhorse. It creates a trial structure that is required by neuroGLM. It sets up the specific glmspike objects for each GLM and fits them. It also has model comparison code to analyze the data.

memming/mtlipglm