Data from: Selective presynaptic inhibition of leg proprioception in behaving Drosophila
Data files
Sep 03, 2025 version files 2.68 GB
-
9A_activation.parquet
337.69 MB
-
9A_silencing.parquet
674.53 MB
-
9A_treadmill_platform.parquet
189.22 MB
-
claw_magnet_Mamiya2018.parquet
849.68 KB
-
claw_treadmill.parquet
144.09 MB
-
club_magnet_Mamiya2018.parquet
389.14 KB
-
club_treadmill_removal.parquet
51.10 MB
-
club_treadmill.parquet
72.88 MB
-
control_activation.parquet
289.35 MB
-
control_silencing.parquet
353.41 MB
-
fanc_dn_information.csv
2.63 KB
-
hook_extension_magnet.parquet
9.27 MB
-
hook_extension_treadmill_platform.parquet
177.69 MB
-
hook_flexion_01_magnet_Mamiya2018.parquet
766.56 KB
-
hook_flexion_01_magnet.parquet
14.99 MB
-
hook_flexion_01_treadmill_platform.parquet
154.53 MB
-
hook_flexion_02_treadmill.parquet
101.74 MB
-
hook_flexion_03_bdn2.parquet
290.44 KB
-
manc_v1_classifications.csv
548.14 KB
-
manc_v1_completeness.csv
281.43 KB
-
manc_v1_connectivity.parquet
13.23 MB
-
pmt_control.parquet
66.44 KB
-
README.md
10.09 KB
-
rna-seq.xlsx
128.69 KB
-
web_treadmill.parquet
94.17 MB
Abstract
Controlling arms and legs requires feedback from proprioceptive sensory neurons that detect joint position and movement. Proprioceptive feedback must be tuned for different behavioral contexts, but the underlying circuit mechanisms remain poorly understood. Using calcium imaging in behaving Drosophila, we find that the axons of position-encoding leg proprioceptors are active across a range of behaviors, whereas the axons of movement-encoding leg proprioceptors are suppressed during walking and grooming. Using connectomics, we identify a specific class of interneurons that provide GABAergic presynaptic inhibition to the axons of movement-encoding proprioceptors. These interneurons receive input from parallel excitatory and inhibitory descending pathways that are positioned to drive the interneurons in a context- and leg-specific manner. Calcium imaging from both the interneurons and their descending inputs confirms that their activity is correlated with self-generated but not passive leg movements. Taken together, our findings provide evidence for a neural circuit that suppresses specific proprioceptive feedback signals during self-generated movements.
This dataset contains the calcium imaging and behavioral data generated in the study, as well as supporting RNA-seq and connectome data.
Dataset DOI: 10.5061/dryad.gqnk98t16
Description of the data and file structure
This dataset contains calcium imaging and behavioral data generated in Dallmann et al. (2025) as well as supporting RNA-seq and connectome data. All data are from Drosophila melanogaster. Calcium imaging data are from behaving animals (on a treadmill, on a platform, or without ground contact) and fully restrained animals (leg movements were externally imposed).
Code to recreate all analyses is provided on GitHub (https://github.com/chrisjdallmann/feco-inhibition).
Calcium imaging data from behaving animals
Files
claw_treadmill.parquet: Calcium imaging data from claw axons during behavior on and off the treadmill.
hook_flexion_01_treadmill_platform.parquet: Calcium imaging data from hook flexion axons (driver line 1) during behavior on and off the treadmill and passive leg movements with the platform.
hook_flexion_02_treadmill.parquet: Calcium imaging data from hook flexion axons (driver line 2) during behavior on and off the treadmill.
hook_extension_treadmill_platform.parquet: Calcium imaging data from hook extension axons during behavior on and off the treadmill and passive leg movements with the platform.
club_treadmill.parquet: Calcium imaging data from club axons during behavior on and off the treadmill.
club_treadmill_removal.parquet: Calcium imaging data from club axons while the treadmill was removed during the recording.
9A_treadmill_platform.parquet: Calcium imaging data from 9A axons during behavior on and off the treadmill and passive leg movements with the platform.
web_treadmill.parquet: Calcium imaging data from DNg74 (web) axons during behavior on the treadmill.
Variables
The parquet files have the following columns:
driver: Name of driver line.
roi: Region of interest imaged in VNC.
animal_id: Animal identifier.
trial: Name of recording.
frame: Frame number. Sampling rate is 300 Hz.
time: Time in seconds.
ball: Presence (1) or absence (0) of the ball treadmill.
platform: Presence (1) or absence (0) of the platform.
analyze: 1 if the frame can be included in the analysis, 0 otherwise. See Dallmann et al. (2025) for criteria.
L1_rest: 1 if the left front leg did not move actively (was at rest), 0 otherwise.
L1_move: 1 if the left front leg moved actively, 0 otherwise.
L1_walk: 1 if the left front leg was walking, 0 otherwise.
L1_groom: 1 if the left front leg was grooming, 0 otherwise.
L1_other: 1 if the left front leg performed movements other than walking or grooming, such as pushing.
annotation: 1 if a frame was manually annotated. See Dallmann et al. (2025) for annotations.
manipulation: 1 if the treadmill was removed during the recording, 0 otherwise.
calcium: Measured changes in intracellular calcium. Normalized, z-scored ratio of GCaMP to tdTomato fluorescence. See Dallmann et al. (2025) for the calculation.
predicted_calcium: Normalized predicted changes in intracellular calcium.
L1C_flex: Flexion of left front leg femur-tibia joint in degrees.
L1C_flex_vel: Flexion velocity of left front leg femur-tibia joint in degrees per second.
vel_forward: Forward velocity of the animal on the treadmill in millimeters per second.
vel_lateral: Lateral velocity of the animal on the treadmill in millimeters per second.
vel_rotational: Rotational velocity of the animal on the treadmill in degrees per second.
Calcium imaging data from fully restrained animals
Files
hook_flexion_01_magnet.parquet: Calcium imaging data from hook flexion axons (driver line 1) during passive leg movements with the magnet.
hook_extension_magnet.parquet: Calcium imaging data from hook extension axons during passive leg movements with the magnet.
claw_magnet_Mamiya2018.parquet: Calcium imaging data from claw axons during passive leg movements with the magnet (Mamiya et al. 2018). Used to tune the computational model.
hook_flexion_01_magnet_Mamiya2018.parquet: Calcium imaging data from hook flexion axons (driver line 1) during passive leg movements with the magnet (Mamiya et al. 2018). Used to tune the computational model.
club_magnet_Mamiya2018.parquet: Calcium imaging data from club axons during passive leg movements with the magnet (Mamiya et al. 2018). Used to tune the computational model.
hook_flexion_03_bdn2.parquet: Calcium imaging data from hook flexion axons with and without activation of descending neuron DNg100 (BDN2).
pmt_control.parquet: GFP fluorescence from FeCO axons following the start-up of the microscope PMT.
Variables
The parquet files have the following columns:
driver: Name of driver line.
roi: Region of interest imaged in VNC.
animal_id: Animal identifier.
trial: Name of recording.
frame: Frame number.
time: Time in seconds.
stimulus_type: Type of stimulus used to passively move the leg. walking_replay, grooming_replay, ramp_hold_flex_first, or ramp_hold_ext_first.
stimulus: 1 if leg was passively moved or if an optogenetic stimulus was presented, 0 otherwise.
stimulus_core: 1 if leg was passively moved with walking- or grooming-like kinematics (excludes transition movements between bouts).
analyze: 1 if frame can be included in the analysis, 0 otherwise. See Dallmann et al. (2025) for criteria.
calcium: Measured changes in intracellular calcium. Normalized, z-scored ratio of GCaMP to tdTomato fluorescence. See Dallmann et al. (2025) for the calculation.
predicted_calcium: Normalized predicted changes in intracellular calcium.
L1C_flex: Flexion of left front leg femur-tibia joint in degrees.
L1C_flex_vel: Flexion velocity of left front leg femur-tibia joint in degrees per second.
Optogenetic activation and silencing data from behaving animals
Files
9A_activation.parquet: Behavioral data for optogenetic activation of 9A neurons in tethered walking flies.
9A_silencing.parquet: Behavioral data for optogenetic silencing of 9A neurons in tethered walking flies.
control_activation.parquet: Behavioral data for optogenetic activation of tethered walking control flies.
control_silencing.parquet: Behavioral data for optogenetic silencing of tethered walking control flies.
Variables
The parquet files have the following columns:
animal_id: Animal identifier.
trial: Name of recording.
frame: Frame number. Sampling rate is 300 Hz.
trial_phase: Phase of trial. Pre-stim indicates the pre-stimulus phase, Stim on indicates the stimulus phase, and Post-stim indicates the post-stimulus phase.
The other columns contain kinematic parameters:
L1, L2, L3, R1, R2, R3: Data of the left (L) or right (R) front (1), middle (2), or hind (3) leg.
A, B, C, D, E: Body-coxa joint (A), coxa-trochanter joint (B), femur-tibia joint (C), tibia-tarsus joint (D), or tip of tarsus (E).
rot: Rotation angle in degrees.
flex: Flexion angle in degrees.
abduct: Abduction angle in degrees.
x, y, z: Coordinates in space.
vel_forward: Forward velocity of the animal on the treadmill in millimeters per second.
vel_lateral: Lateral velocity of the animal on the treadmill in millimeters per second.
vel_rotational: Rotational velocity of the animal on the treadmill in degrees per second.
*_bout_step_num: Number of steps in recording.
*_phase: Step cycle phase in radians.
RNA-seq data
Files
rna-seq.xlsx: RNA-seq data for claw, hook, and club neurons.
Variables
The xlsx file has three sheets, one for each type of neuron (claw, hook, and club). Each sheet has the following columns:
cellID: Cell identifier.
Rdl, Lcch3, Grd, GABA-B-R1, GABA-B-R2, GABA-B-R3, nAChRalpha1, nAChRalpha3, nAChRalpha4, nAChRalpha5, nAChRalpha6, nAChRalpha7, nAChRbeta1, nAChRbeta2, nAChRbeta3, mAChR-A, mAChR-B, mAChR-C, GluRIB, GluRIIA, GluRIIC, GluRIID, GluRIIE, GluClalpha, mGluR, Nmdar1, Nmdar2: Expression level of receptor gene.
Connectome data
Files
fanc_dn_information.csv: Connectome identifiers, predicted neurotransmitters, and community names for descending neurons of interest in the FANC connectome.
manc_v1_classifications.csv: Classification of neurons of the MANC connectome (version 1.0).
manc_v1_completeness.csv: Connectome identifiers for neurons of the MANC connectome (version 1.0) that can be included in the simulation.
manc_v1_connectivity.parquet: Weighted synaptic connectivity between neurons of the MANC connectome (version 1.0).
Variables
fanc_dn_information.csv has the following columns:
pt_root_id: Connectome identifier of neuron.
nt_type: Predicted neurotransmitter of neuron. GLUT indicates glutamate, ACH indicates acetylcholine, nan indicates that a neurotransmitter prediction is not available.
name: Community name of neuron. nan indicates that a community name is not available.
manc_v1_classifications.csv has the following columns:
bodyId: Connectome identifier of neuron.
class: Classification of neuron (e.g., descending neuron). nan indicates that a classification is not available.
manc_v1_completeness.csv has the following columns:
bodyId: Connectome identifier of neuron.
Completed: TRUE if the neuron has been proofread and can be included in the connectome simulation.
manc_v1_connectivity.parquet has the following columns:
Presynaptic_ID: Connectome identifier of presynaptic neuron.
Postsynaptic_ID: Connectome identifier of postsynaptic neuron.
Presynaptic_Index: Index of presynaptic neuron.
Postsynaptic_Index: Index of postsynaptic neuron.
Excitatory x Connectivity: Weighted synaptic connectivity between pre- and postsynaptic neurons. Positive values indicate an excitatory connection, negative values indicate an inhibitory connection.