Data associated with "A weakened recurrent circuit in the hippocampus of Rett syndrome mice disrupts long-term memory representations"

Datasets used in A weakened recurrent circuit in the hippocampus of Rett syndrome mice disrupts long-term memory representations.

Datatypes:

Multi-index pandas dataframe (.pkl)

Numpy array (.npy)

Collection of numpy arrays (.npz)

Python dictionary objects (.pkl)

Datasets:

alignments.pkl: A dataframe containing numpy arrays of image displacements for each mouse in each memory context.

This multi-index dataframe has rows indexed by genotype ('wt' or 'het') and mouse_id. The columns are ['T', 'F1', 'N1', 'F2', 'N2'] for the training, recall 1-hour, neutral, recall 1-day, neutral day 2 memory contexts respectively. Each element of this dataframe is a numpy array of shape  images x 2 that hold x and y image displacements respectively. These alignments are computed after the inscopix software motion correction and are used in Supplemental Figure 2 of the paper.

behavior_df.pkl: A dataframe of behavior readouts recorded by a camera positioned above the mice in each context chamber.

This multi-index dataframe has rows indexed by genotype ('wt' or 'het') and mouse_id.The columns are; sample times (*_time), freezing boolean arrays (*_freeze), x-positions in context chamber (*_x) and y-positions in the context chamber (*_y) for each context (*) in ('Train', 'Fear', 'Neutral', 'Fear_2', 'Neutral_2').

correlated_pairs_df.pkl: A dataframe containing arrays of neuron indices that have a correlation in activity pattern > 0.3.

This multi-index dataframe has rows indexed by genotype ('wt' or 'het') and mouse_id and treatment ('NA'). The columns contain ['Train', 'Fear', 'Neutral', 'Fear_2', 'Neutral_2'] representing each memory context. Each element of the dataframe is a numpy array with three columns. The first two columns are the neuron indices that are correlated and the last column is the strength of the correlation.

dredd_freezes_df.pkl: A dataframe containing freezing percentages for SOM-Cre and RTT-SOM-Cre mice treated with DREADDS.

This multi-index dataframe has rows indexed by genotype ('wt' or 'het') and mouse_id and treatment (mcherry, hm3d, hm4d). The columns contain one of ['Neutral', 'Fear', 'Fear_2']. Each element of the dataframe is a freezing percentage for a single mouse. This dataframe is built from reading the dredd_behavior.xlsx excel file. This is used to generate figure 5E of the paper.

high_degree_df.pkl: A dataframe containing list of high degree neuron indices.

This multi-index dataframe has rows indexed by genotype ('wt' or 'het') and mouse_id and treatment ('NA'=not applicable since no DREADD used). The columns contain ['Train', 'Fear', 'Neutral', 'Fear_2', 'Neutral_2'] representing each memory context. Each element of the dataframe is a list of neuron indices that are high-degree cells.

N006_wt_basis.npz: a dict containing three numpy arrays representing the basis images for mouse N006 of genotype wild-type.

This dict has three arrays stored under the variable names 'U', 'sigma' and 'img_shape'. U is a matrix of column vector basis images. Each column is the vector representation of a basis image (row pixels x column pixels). There are 220 basis images (columns) in U. The sigma variable is the singular value associated with each basis image vector in U. img_shape can be used to reshape each basis column vector into a 2-D image for viewing. This data is used in Supplemental Figure 2 of the paper.

N006_wt_cxtbasis.pkl: A dictionary containing arrays for basis images and singular values for each context.

This dictionary has keys, ['Train', 'Fear', 'Neutral', 'Fear_2',  'Neutral_2'] representing the memory contexts. Each value is a 2 element list containing the U-basis images as column vectors and singular values, one per basis image in U. The shape of the basis images is the same shape stored in N006_wt_basis.pkl. This dataset is used in Supplementary Figure 2 to track cells across contexts of the CFC task (see also N006_wt_cxtsources.pkl)

N006_wt_cxtsources.pkl: A dictionary containing the independent component source images computed from the basis images for automatically identifying regions of interest (ROIs). 

The dictionary is keyed on  ['Train', 'Fear', 'Neutral', 'Fear_2',  'Neutral_2'] contexts. Each value in the dictionary at a given key is a 3-D numpy array of shape sources x height x width. These data were used to construct the source images and max intensity projection image of the sources in Supplemental Figure 2F-J of the paper.

N006_wt_rois.pkl: A dictionary containing the boundaries and annuli coordinates of all rois for mouse N006 of genotype wild-type.

This dictionary is keyed on ['boundaries', 'annuli'] contexts and each value is a 179 element list of arrays of boundary line coordinates or annulus point coordinates one  per ROI detected for this mouse.

N006_wt_sources.npy: A numpy array containing all source images computed from all contexts of the CFC task for mouse N006 of genotype wild-type.

This numpy array has shape n x height x width where n=205 source images, height=517 pixels and width=704 pixels. This data was used to construct Supplemental Figure 3F.

N019_wt_basis.npz: a dict containing three numpy arrays representing the basis images for mouse N019 of genotype wild-type.

This dict has three arrays stored under the variable names 'U', 'sigma' and 'img_shape'. U is a matrix of column vector basis images. Each column is the vector representation of a basis image (row pixels x column pixels). There are 220 basis images (columns) in U. The sigma variable is the singular value associated with each basis image vector in U. img_shape can be used to reshape each basis column vector into a 2-D image for viewing. This data is used in Figure 1C of the paper.

N019_wt_sources.npy: A numpy array containing all source images computed from all contexts of the CFC task for mouse N019 of genotype wild-type.

This numpy array has shape n x height x width where n=204 source images, height=516 pixels and width=698 pixels. This data was used to construct Figure 1C of the paper.

P80_animals.pkl: A pandas multi-index object containing the genotype, mouse_id and treatment of the top 80% behavioral performance animals.

In this study, we drop the lowest 20% performing WT and RTT animals based on freezing percentage during the recall contexts. This multi-index is used to filter the data before each computation or plot in this study. So for example Figure 1B contains only the top 80% performing WT and RTT mice.

pc_sipscs_amps.pkl: A dictionary containing the amplitudes of spontaneous IPSCs recorded in pyramidal cells of WT and RTT mice.

This dictionary is keyed on ['wt', 'mecp2_pos', 'mecp2_neg'] representing whether the pyramidal cell was recorded from a wild-type mouse ('wt') or is an MeCP2 negative or MeCP2 positive RTT cell. This value under each key is an array of IPSC amplitudes, one per recorded cell. This data was used to construct Figure 4C in the paper.

pc_sipscs_freqs.pkl: A dictionary containing the frequencies of spontaneous IPSCs recorded in pyramidal cells of WT and RTT mice.

This dictionary is keyed on ['wt', 'mecp2_pos', 'mecp2_neg'] representing whether the pyramidal cell was recorded from a wild-type mouse ('wt') or is an MeCP2 negative or MeCP2 positive RTT cell. This value under each key is an array of IPSC frequencies, one per recorded cell. This data was used to construct Figure 4C in the paper.

rois_df.pkl: A multi-index dataframe containing all ROI information for each non-DREADD treated cell in this study (Figures 1-3).

This dataframe index contains the genotype ('wt', 'het'), the mouse_id, the treatment ('NA'=not applicable since no DREADD used), and the cell index starting from 0. The columns are ['centroid', 'cell_boundary', 'annulus_boundary']. The centroid for each cell is a 2-tuple of row, column pixel centroid coordinates. The cell_boundary is a two-column array of row, col boundary points for each ROI. The annulus_boundary is a two-column array of row, column interior points in the annulus. The annulus region  excludes points of overlap with nearby cell bodies (See STAR methods of the paper).

signals_df.pkl: A multi-index dataframe containing calcium signals, inferred spikes and metadata for all Non-DREADD experiments used in this study (Figs 1-3).

This dataframe index contains the genotype ('wt', 'het'), the mouse_id, the treatment ('NA'=not applicable since no DREADD used), and the cell index starting from 0 and going up to 5771 cells. The columns are ['channels', 'channel', 'num_pages', 'width', 'height', 'bits', 'Train_signals', 'Fear_signals', 'Neutral_signals', 'Cue_signals', 'Fear_2_signals', 'Neutral_2_signals', 'Cue_2_signals', 'Train_spikes', 'Fear_spikes', 'Neutral_spikes', 'Cue_spikes', 'Fear_2_spikes', 'Neutral_2_spikes', 'Cue_2_spikes', 'sample_rate']. The channels are all the recorded channels, the channels is the channel on which ROIs were detected, the width and height are the image dimensions, the bits is the image bit depth of the calcium movie. The *_signals' are the df/f signals for each cell in each context. Each signal is a numpy array with the first 800 samples have been set to NAN due to settling time of the miniscope. The '*_spikes' are the inferred spikes for each cell stored as an image index. This signal and spike indices can be converted to time using the sample column. This dataframe is used in the construction of Figures 1-3 in the paper.

som_behavior_df.pkl: A dataframe of behavior readouts recorded by a camera positioned above the mice in each context chamber.

This multi-index dataframe has rows indexed by genotype ('wt' or 'het') and mouse_id. The columns are; sample times (*_time), freezing boolean arrays (*_freeze), x-positions in context chamber (*_x) and y-positions in the context chamber (*_y) for each context in *=('Train', 'Fear', 'Neutral', 'Fear_2', 'Neutral_2'). This dataframe was not used in the paper but may still be useful for further analysis.

som_sepsc_amplitudes: A dictionary containing the amplitudes of spontaneous EPSCs recorded in SOM cells of WT and RTT mice with and without MeCP2.

A dictionary with keys ['som', 'som_rett_pos', 'som_rett_neg'] for WT SOM and RTT-SOM cells with and without MeCP2 respectively. Each value is a list of sEPSC amplitudes. This data was used in Figure 4E-G.

som_sepsc_freqs: A dictionary containing the amplitudes of spontaneous EPSCs recorded in SOM cells of WT and RTT mice with and without MeCP2.

A dictionary with keys ['som', 'som_rett_pos', 'som_rett_neg'] for WT SOM and RTT-SOM cells with and without MeCP2 respectively. Each value is a list of sEPSC frequencies. This data was used in Figure 4E-G.

som_signals_df.pkl: A multi-index dataframe containing calcium signals, inferred spikes and metadata for all Non-DREADD SOM cell recordings used in this study (Figs 5).

This dataframe index contains the genotype ('wt', 'het'), the mouse_id, the treatment ('NA'=not applicable since no DREADD used), and the cell index starting from 0 and going up to 710 cells. The columns are ['channels', 'channel', 'num_pages', 'width', 'height', 'bits', 'Train_signals', 'Fear_signals', 'Neutral_signals', 'Cue_signals', 'Fear_2_signals', 'Neutral_2_signals', 'Cue_2_signals', 'Train_spikes', 'Fear_spikes', 'Neutral_spikes', 'Cue_spikes', 'Fear_2_spikes', 'Neutral_2_spikes', 'Cue_2_spikes', 'sample_rate']. The channels are all the recorded channels, the channels is the channel on which ROIs were detected, the width and height are the image dimensions, the bits is the image bit depth of the calcium movie. The *_signals' are the df/f signals for each cell in each context. Each signal is a numpy array with the first 800 samples have been set to NAN due to settling time of the miniscope. The '*_spikes' are the inferred spikes for each cell stored as an image index. This signal and spike indices can be converted to time using the sample column. This data was used to construct Figure 5B-C.

ssn33_sstcre_basis.npz: a dict containing three numpy arrays representing the basis images for mouse ssn33 of genotype sst-cre.

This dict has three arrays stored under the variable names 'U', 'sigma' and 'img_shape'. U is a matrix of column vector basis images. Each column is the vector representation of a basis image (row pixels x column pixels). There are 100 basis images (columns) in U. The sigma variable is the singular value associated with each basis image vector in U. img_shape can be used to reshape each basis column vector into a 2-D image for viewing. This data is used in Figure 5A of the paper.

ssn33_sstcre_sources.npy: A numpy array containing all source images computed from all contexts of the CFC task for mouse N019 of genotype wild-type.

This numpy array has shape n x height x width where n=86 source images, height=516 pixels and width=654 pixels. This data was used to construct Figure 5A of the paper.