Study

Curated
Sato et al., 2025
Neurogenomic and behavioral principles shape freezing dynamics and synergistic performance in Drosophila melanogaster
[10.1038/s41467-025-61313-z]

Description

Flies of 104 DGRP lines (both sexes) were video-tracked alone or in single-sex groups of six, before and during repeated looming stimuli: locomotor speed, freezing duration, nearest-neighbour distance, male chasing and visual responsiveness to conspecifics. GWAS highlighted Ptp99A and kirre.

Comments from curator

Phenotyping data comes from the authors' figshare archive (10.6084/m9.figshare.25658730, Sato2025_Fruitfly_collectives.tar.gz, folder data/1_single_strain/parquet), built by their script 1_1_make_dataset.R (code on Zenodo and GitHub):
- df_f5min_speed: mean speed of every fly in the first 5 min.
- df_s5min_2995_5990_freezing_duration: freezing duration of every arena.
- df_f5min_nnd_rand: nearest-neighbour distance (NND) at rest of every group arena.
- df_motion_cue_exit_coeff_trial and df_motion_cue_exit_coeff: visual responsiveness, per group arena and per line.
- df_f5min_group_male_seconds_nest_dist_chase_3mm_strain_mean: chasing and chaining durations of every male group arena.
and from Supplementary Table 2 (41597_2025_4724_MOESM2_ESM.xlsx) of the companion data descriptor of the same experiment (Sato, Okuyama & Takahashi 2025, Sci Data 12:400): per-line mean and SD of the time-averaged NND and of the time spent in the arena centre. The Source Data file of the paper (41467_2025_61313_MOESM14_ESM.xlsx), with the per-line means and SE of Figures S1-S5, was used to check the uploads; its sheet FigureS7b gives the speed-recovery parameters of 15 lines.

Design: 104 DGRP lines; flies 2-4 days old (25 °C, 12:12 h light:dark, Bloomington food) were tested alone or in single-sex groups of 6 in 30-mm arenas (12 arenas per plate) under a monitor, for a 10-min session: 5 min without stimulus, then 20 looming stimuli (500 ms, every 15 s) mimicking a predator. Videos (about 50 frames/s) were tracked with Tracktor and the positions averaged in 0.5-s bins. 20 arenas were recorded per line, sex and context (single or group), between October 2021 and February 2022. The recordings last 15 min; the paper analyses the first 10 (the companion Sci Data paper describes the whole session).

raw_1 has one row per arena: row k of a line and sex holds the k-th single-fly arena and the k-th group arena, which are independent (the arena ids are kept as obsolete design columns ArenaID_Single and ArenaID_Group). Its phenotypes:
- Speed_Single (mm/s): mean moving speed of the single fly in the first 5 min.
- FreezeDur_Single and FreezeDur_Group (s): first time after the stimulus (0-14 s, 0.5-s steps) at which the speed of the fly, averaged over the 20 stimuli and divided by its mean speed of the first 5 min, is back above 1. Flies that never recover within 14 s get 14, and this coded value is kept, as in the authors' means, figures and GWAS (1231 of the 4160 single flies, 21 of the 4160 group arenas). For groups, the value of the arena is the mean of its 6 flies: the published files have no per-fly freezing durations.
- NND_Group (mm): mean distance of the 6 flies to their nearest neighbour at the time point of the first 5 min where they move least (at rest), as in the Methods and Figure S2.
- VisResp_b0 and VisResp_b1: intercept and slope of the logistic regression of freezing exit (a fly stopped 0.5 s after a stimulus walks > 4 mm/s at 10 s) on the increase of the motion cue it perceives from the other 5 flies, fitted by the authors for each group arena. The intercept is the paper's "visual responsiveness" (Figure S5b). These per-arena fits are unstable for some arenas (|beta0| > 20 in 168 of 4160 arenas), so the line medians are more robust than the means. The slope is in the authors' files but not used in the paper.
- ChaseTime and ChainTime (s): chasing (both flies moving > 1 mm/s, headings within pi/4, distance < 3 mm) and chaining (chains of 3 or more chasing flies) durations in the first 5 min, for male groups only (NA for females).
raw_2 has one row per fly of the group arenas (6 flies x 20 arenas per line and sex): Speed_Group (mm/s), with its arena (ArenaID, obsolete).
The summary dataset has:
- mn_VisRespPooled_b0 (and its slope mn_VisRespPooled_b1): the logistic regression fitted once on the pooled flies, stimuli and arenas of the line and sex, which is the visual-responsiveness phenotype of the authors' GWAS (figshare df_motion_cue_exit_coeff). It is not the mean of the per-arena intercepts (Spearman 0.95), and it cannot be derived from them.
- mn_NND5min_Group / sd_NND5min_Group (mm): NND averaged over the first 5 min, and mn_CenterTime_Single / sd_CenterTime_Single and mn_CenterTime_Group / sd_CenterTime_Group (s): time spent within 7.5 mm of the arena centre in the first 5 min ("boldness"), from the Sci Data Supplementary Table 2. The centre time is not in the Nat Commun paper.
- mn_SpdRecov_a_Group and mn_SpdRecov_b_Group: parameters a and b of the authors' fit S(t) = a ln(t) + b of the speed recovery after the looming stimuli (S = speed t s after a stimulus relative to the speed 0.5 s before the stimuli, t = 0.5-14 s), for 15 lines and females only: they come from the single-strain groups of 6 females recorded as controls in the mixed-strain experiment (same assay), and are copied from the Source Data of Figure S7b, where they parametrise the simulations.

The control strains were dropped: the visually impaired mutant w* norpAP12 (BDSC 9047), norpAP12 backcrossed four times into DGRP_208 (DGRP208_norpA), and the "random" pseudo-groups of Figure S2 (random combinations of single flies). All 104 lines match the BDSC numbers of the strain list (Sci Data Supplementary Table 1). Sexes were tested separately (F and M).

Checks: the means and SE computed from the raw datasets reproduce the Source Data of Figures S1 (speed, single and group), S2 (NND), S3 (chasing and chaining), S4 (freezing duration) and S5b (visual responsiveness) for all 104 lines and both sexes (differences < 1e-13), and the means and SD of speed and freezing duration of the Sci Data Supplementary Table 2 (to its 9 decimals). The line means of FreezeDur_Single, FreezeDur_Group and mn_VisRespPooled_b0 rank the lines as the authors' GWAS phenotype files (figshare gwas/pheno, bestNormalize-transformed line means; Spearman 0.9999 to 1).

Of note, the NND used for the GWAS is the time-averaged NND, not the NND at rest: the authors' NND GWAS phenotype files (dated January 2024) rank the lines exactly as mn_NND5min_Group (Spearman 1.0 in each sex), whereas the NND at rest described in the Methods and shown in Figure S2 (NND_Group, files of September 2024) correlates only at 0.46 (females) and 0.61 (males). Both are uploaded; mn_NND5min_Group is the one to use to reproduce the paper's NND GWAS. Also, the group values of the centre time exceed 300 s (the length of the period) for 36 of the 208 line x sex (up to 466.65 s), so they are not the time of one fly; the data descriptor does not say how the 6 flies were combined (probably summed).

The authors' GWAS (GCTA fastGWA-mlm with 10 genotype PCs, 1,689,110 SNPs with MAF > 0.05, each sex separately, on bestNormalize-transformed line means) used the freezing duration (single and group), the NND and the visual responsiveness; Ptp99A and kirre were associated with visual responsiveness in both sexes. Broad-sense heritabilities: speed 0.94-0.96, freezing duration 0.68-0.84, NND 0.40 (F) and 0.53 (M), visual responsiveness 0.53 (F) and 0.51 (M).

Not uploaded:
- The per-fly and per-stimulus events of Figures 1d and S5a (inputs of the logistic regressions), and the alternative visual-responsiveness metric of Figure S20 (motion cue accumulated until freezing exit), which is published only per fly and stimulus (figshare parquet files) and as transformed GWAS values.
- The mixed-strain experiment (groups of 3 + 3 females from 2 of 15 DGRP lines, 105 pairs): group-level values of pairs of lines, and the speed-recovery time courses of its single-strain control groups (Figure S7a; only their fitted parameters a and b are uploaded).
- The GAL4/UAS experiments (lamina neurons and Ptp99A; not DGRP lines), the spider animal-computer-interaction experiments, the simulations, and the reanalysis of single-cell RNA-seq of visual neurons (GSE156455, previously published).
The raw videos and trajectories of all flies are on Dryad (10.5061/dryad.ttdz08m60), described in the Sci Data paper. 1552 of the single-fly recordings of this experiment are reused, as reference single flies, by Okuyama, Sato & Takahashi 2026 (iScience, DGRPool study 252).

Other external data available

10.6084/m9.figshare.25658730 figshare archive of the paper (behavioural, GWAS and simulation datasets; source of the per-arena and per-fly data uploaded here)
10.5061/dryad.ttdz08m60 Dryad deposit with the raw videos and tracking trajectories of all flies (companion Sci Data paper)
10.5281/zenodo.15630079 analysis code of the paper
GSE156455 single-cell RNA-seq of developing visual neurons (previously published, reanalysed in the paper)

Datasets

Dataset Phenotypes DGRP lines
Summary dataset 10 phenotypes
mn_CenterTime_Group mn_CenterTime_Single mn_NND5min_Group mn_SpdRecov_a_Group mn_SpdRecov_b_Group mn_VisRespPooled_b0 mn_VisRespPooled_b1 sd_CenterTime_Group sd_CenterTime_Single sd_NND5min_Group
104 DGRP lines
DGRP_021
DGRP_026
DGRP_028
DGRP_041
DGRP_045
DGRP_057
DGRP_069
DGRP_073
DGRP_075
DGRP_088
DGRP_093
DGRP_101
DGRP_105
DGRP_109
DGRP_129
DGRP_136
DGRP_142
DGRP_149
DGRP_153
DGRP_161
DGRP_176
DGRP_181
DGRP_189
DGRP_195
DGRP_208
DGRP_217
DGRP_228
DGRP_235
DGRP_280
DGRP_287
DGRP_301
DGRP_303
DGRP_309
DGRP_315
DGRP_318
DGRP_320
DGRP_324
DGRP_335
DGRP_340
DGRP_357
DGRP_358
DGRP_360
DGRP_362
DGRP_365
DGRP_370
DGRP_373
DGRP_375
DGRP_379
DGRP_382
DGRP_385
DGRP_399
DGRP_405
DGRP_427
DGRP_437
DGRP_443
DGRP_461
DGRP_486
DGRP_502
DGRP_508
DGRP_513
DGRP_517
DGRP_531
DGRP_535
DGRP_555
DGRP_563
DGRP_584
DGRP_589
DGRP_703
DGRP_707
DGRP_712
DGRP_716
DGRP_721
DGRP_730
DGRP_732
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_774
DGRP_786
DGRP_796
DGRP_799
DGRP_801
DGRP_804
DGRP_808
DGRP_810
DGRP_820
DGRP_822
DGRP_832
DGRP_837
DGRP_843
DGRP_849
DGRP_852
DGRP_855
DGRP_857
DGRP_859
DGRP_879
DGRP_884
DGRP_890
DGRP_892
DGRP_897
DGRP_907
DGRP_911
DGRP_913
Download
Raw dataset #1 8 phenotypes
ChainTime ChaseTime FreezeDur_Group FreezeDur_Single NND_Group Speed_Single VisResp_b0 VisResp_b1
104 DGRP lines
DGRP_021
DGRP_026
DGRP_028
DGRP_041
DGRP_045
DGRP_057
DGRP_069
DGRP_073
DGRP_075
DGRP_088
DGRP_093
DGRP_101
DGRP_105
DGRP_109
DGRP_129
DGRP_136
DGRP_142
DGRP_149
DGRP_153
DGRP_161
DGRP_176
DGRP_181
DGRP_189
DGRP_195
DGRP_208
DGRP_217
DGRP_228
DGRP_235
DGRP_280
DGRP_287
DGRP_301
DGRP_303
DGRP_309
DGRP_315
DGRP_318
DGRP_320
DGRP_324
DGRP_335
DGRP_340
DGRP_357
DGRP_358
DGRP_360
DGRP_362
DGRP_365
DGRP_370
DGRP_373
DGRP_375
DGRP_379
DGRP_382
DGRP_385
DGRP_399
DGRP_405
DGRP_427
DGRP_437
DGRP_443
DGRP_461
DGRP_486
DGRP_502
DGRP_508
DGRP_513
DGRP_517
DGRP_531
DGRP_535
DGRP_555
DGRP_563
DGRP_584
DGRP_589
DGRP_703
DGRP_707
DGRP_712
DGRP_716
DGRP_721
DGRP_730
DGRP_732
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_774
DGRP_786
DGRP_796
DGRP_799
DGRP_801
DGRP_804
DGRP_808
DGRP_810
DGRP_820
DGRP_822
DGRP_832
DGRP_837
DGRP_843
DGRP_849
DGRP_852
DGRP_855
DGRP_857
DGRP_859
DGRP_879
DGRP_884
DGRP_890
DGRP_892
DGRP_897
DGRP_907
DGRP_911
DGRP_913
Download
Raw dataset #2 1 phenotypes
Speed_Group
104 DGRP lines
DGRP_021
DGRP_026
DGRP_028
DGRP_041
DGRP_045
DGRP_057
DGRP_069
DGRP_073
DGRP_075
DGRP_088
DGRP_093
DGRP_101
DGRP_105
DGRP_109
DGRP_129
DGRP_136
DGRP_142
DGRP_149
DGRP_153
DGRP_161
DGRP_176
DGRP_181
DGRP_189
DGRP_195
DGRP_208
DGRP_217
DGRP_228
DGRP_235
DGRP_280
DGRP_287
DGRP_301
DGRP_303
DGRP_309
DGRP_315
DGRP_318
DGRP_320
DGRP_324
DGRP_335
DGRP_340
DGRP_357
DGRP_358
DGRP_360
DGRP_362
DGRP_365
DGRP_370
DGRP_373
DGRP_375
DGRP_379
DGRP_382
DGRP_385
DGRP_399
DGRP_405
DGRP_427
DGRP_437
DGRP_443
DGRP_461
DGRP_486
DGRP_502
DGRP_508
DGRP_513
DGRP_517
DGRP_531
DGRP_535
DGRP_555
DGRP_563
DGRP_584
DGRP_589
DGRP_703
DGRP_707
DGRP_712
DGRP_716
DGRP_721
DGRP_730
DGRP_732
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_774
DGRP_786
DGRP_796
DGRP_799
DGRP_801
DGRP_804
DGRP_808
DGRP_810
DGRP_820
DGRP_822
DGRP_832
DGRP_837
DGRP_843
DGRP_849
DGRP_852
DGRP_855
DGRP_857
DGRP_859
DGRP_879
DGRP_884
DGRP_890
DGRP_892
DGRP_897
DGRP_907
DGRP_911
DGRP_913
Download
Back to studies