Companion Data Page Individuality in Generalized Learning Responses Matthew AY Smith1, Kyle Honneger2, Glenn Turner3, Benjamin de Bivort1 1 - Department of Organismic and Evolutionary Biology and Center for Brain Science, Harvard University, Cambridge, Massachusetts, USA. 2 - Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, Illinois, USA. 3 - Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, VA, USA. Manuscript PDF
Download experimental data and analysis scripts (Zenodo)
Zenodo repository	Archive .zip file containing all experimental data and analysis scripts.
Olfactory Learning Behavior data
fig1_flyData.mat	A .mat file that contains a structure array containing behavior data for 75 Canton-S flies within the olfactory learning rig. There are 75 instances with two fields, forwardLrn and reverseLrn. Each instance represents a unique fly and the associated data for the forward and reversal learning assays. Within each learning trial there is "centroid", "shock", "realTime", and "Stimulus". Centroid represents the x,y coordinates of the fly within the odor tunnel. The shock field is a binary array of True, if shock is applied, and False, if shock is not applied for each time point. The stimulus field indicates what odors are being presented, concentration, and time applied. The realTime contains an array of time in seconds for each timepoint within the experiment.
fig2ab_flyData.mat	A .mat file that contains a structure array containing behavior data for 75 flies expressing CsChrimson in bitter taste receptor neurons behaving within the olfactory learning rig. There are 75 instances with four fields representing the CS+ of the four learning trials: pentanol, heptanone, MCH, and octanol. Each instance represents a unique fly and the associated data for the forward and reversal learning assays. Within each learning trial there is "centroid", "shock", "realTime", and "Stimulus". Centroid represents the x,y coordinates of the fly within the odor tunnel. The shock field is a binary array of True, if shock is applied, and False, if shock is not applied for each time point. The stimulus field indicates what odors are being presented, concentration, and time applied. The realTime contains an array of time in seconds for each timepoint within the experiment.
fig2cd_flyData.mat	A .mat file that contains a structure array containing behavior data for 60 flies expressing CsChrimson in bitter taste receptor neurons (gr66a) behaving within the olfactory learning rig. There are 60 instances with two fields representing the forward and reverse learning assays using optogenetics and shock. Within each learning trial there is "centroid", "shock", "realTime", and "Stimulus". Centroid represents the x,y coordinates of the fly within the odor tunnel. The shock field is a binary array of True, if shock is applied, and False, if shock is not applied for each time point. The stimulus field indicates what odors are being presented, concentration, and time applied. The realTime contains an array of time in seconds for each timepoint within the experiment.
Scripts for behavior data
lrnProtocol_updated.m	A MATLAB script that takes the behavior data as an input and generates a figure displaying the unconditioned and conditioned stimulus presentation as square waves, indictating when a stimulus is on or off.
tracksPlot2020.m	A MATLAB script that takes the behavior data as an input and generates a figure displaying the centroid position of flies within the odor tunnels. The background is colored based on the odor or air presentation at a given time.
lrnScoreUpdate.m	A MATLAB script that takes the behavior data as an input and generates the learning score for each individual. It detects which odor is the CS+ through matching the US triggering with the presentation of a particular odor.
figure1_updated.m	A MATLAB script that takes the behavior data from fig1_flyData.mat and performs the behavioral analysis and generates the figures presented in Figure 1 of the paper.
figure2ab_updated.m	A MATLAB script that takes the behavior data from fig2ab_flyData.mat and performs the behavioral analysis and generates the figures presented in panels A and B of Figure 2 of the paper.
figure2cd_updated.m	A MATLAB script that takes the behavior data from fig2cd_flyData.mat and performs the behavioral analysis and generates the figures presented in panels C and D of Figure 2 of the paper.
mattbar_norm.m	A MATLAB script that generates bar graphs with bootstrapped mean and std. The function takes a cell array as input, with each cell containing an array of elements to be included in the bar graph.
lrnGraph.m	A MATLAB script that will generate a plot displaying the change in octanol occupancy for each fly in the passed flyStruct data structure. The function takes in pre-training occupancy score and post-training occupancy score. An example of use in figure1_updated.m.
mattCorr.m	A MATLAB script that will generate a scatter plot with Pearson-r and p values overlayed onto the plot. It takes two equal sized vectors as the input.
cleanFormat.m	A MATLAB script for formatting figures.
PCARegressionCI.m	Adapted from https://github.com/de-Bivort-Lab/dblab-utility-functions/blob/master/linearRegressionCI.m ;performs linear regression on the two column variables of data using PCA to find the regression fit, i.e., using normals to compute residuals performs bootstrap resampling of this regression to determine the confidence interval of the regression line.