DANDI:000467

the breadth of glomerular responses varied dramatically across odorants, with most odorants activating a few glomeruli, while 3-hexanone activated large regions of the ventral/medial AL

Results, Recording calcium responses to general odorants, paragraph 2; Figure 3F-G

3-hexanone has the largest area of at least 0.2 in its trial-averaged max projection in all 6 ants; median areas are 2,810 um2 for 3-hexanone, 1,092 for ethylpyrazine, 390 for isopropanol, 342 for propionic acid and 54 for ethanol. The ratio to ethylpyrazine is 2.57 (2.58 without the AL mask), to the others 7.2 to 51.6. In round 1, with a 13-pixel Minimum filter in place of ImageJ's 21-pixel radius-2 kernel, the ratio was 1.91 and the outcome partially reproduced; the round-1 areas are not shown by the current notebook.

How the test differed from the paper’s method

• Area counted inside the AL mask; the whole field of view is shown as a variant and gives the same.
• The Minimum filter uses ImageJ's radius-2 kernel (21 pixels, RankFilters); round 1 used a 13-pixel disc.

we first imaged responses to 4-methyl-3-hexanol in both ALs simultaneously (n=3 ants) and found that response patterns were bilaterally symmetrical

Results, Pheromone representation is sparse, paragraph 1; Figure S4A

Both ALs have responding ROIs in all three ants under every reading, and the mirrored maps of responding pixels correlate at r = 0.143, 0.185 and 0.072 against -0.015, -0.024 and -0.008 for the half-width shift. The ROI counts differ by more than one in at least one ant under every reading (bix 4 and 5, 4 and 8, 1 and 3; aix 5 and 8, 4 and 5, 3 and 4). One of the criterion's two parts holds in every ant, so the outcome is partial. The verifier would give not reproduced, since the criterion needs both parts and defines no partial outcome; the disagreement is recorded under verifications.

How the test differed from the paper’s method

• The two ALs lie one above the other in the 1X images; the midline is the least fluorescent row between 30% and 70% of the height, and the files do not say which AL is the left one.
• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.

each odorant activated a unique combination of glomeruli, and responses to the same odorant occurred in similar regions of the AL in different individuals

Results, Recording calcium responses to general odorants, paragraph 2; Figure 3F

Across ants, 10 of 27 ant and odorant cases have their centroid closer to the same odorant in the other ants than to any other odorant (3-hexanone in 5 of its 6 ants; ethanol has no pixel of at least 0.2 in one ant, so 27 cases); the criterion needs more than half, and the centroid test does not depend on the M2 reading. Within ants, identical odorant pairs occur under every M2 reading (1 to 9 summed over the 6 ants; bix 7).

How the test differed from the paper’s method

• Stabilization by whole-pixel phase correlation in place of Fiji's Image Stabilizer; ROIs drawn by rule inside an AL mask (log-Otsu on baseline fluorescence), not by eye.
• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.
• The Minimum filter uses ImageJ's radius-2 kernel (21 pixels, RankFilters); round 1 used a 13-pixel disc.
• "Closer" was read as a smaller mean distance to the same odorant's centroids in the other ants than to each other odorant's centroids; the claims file does not fix the distance rule.

Even at the highest concentration tested, the four pheromones activated a median of at most 6 glomeruli

Results, Pheromone representation is sparse, paragraph 2; Figure S5A; Figure 6 legend

Under bix the medians over ants are 1 (heptanone), 1 or 0 (heptanol), 3 (hexanol) and 0 (6MHO) at 48% (8 ants) and at each ant's higher concentration (13 ants); the largest is 3. The four readings b and aix keep every median at or below 6.6 (largest 3 to 6); aiy (7.8 and 8.0), aiix and aiiy (13 to 14.3) do not. Leaving out the one flagged trial changes nothing.

How the test differed from the paper’s method

• The paper drew ROIs by eye. The readings disagree, and the outcome rests on the ROI's mean value (ImageJ's default measure, value i) with the Methods' rule of the trial-mean value (counting x), under which both drawings (aix and bix) meet the criterion, adopted as an assumption; bix is the adopted reading throughout. All readings are shown.
• Stabilization by whole-pixel phase correlation in place of Fiji's Image Stabilizer; ROIs drawn by rule inside an AL mask (log-Otsu on baseline fluorescence), not by eye.
• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.
• Overlapping ROIs of two odors were taken as one glomerulus when at least half of the smaller overlaps the larger and their depths are within 2 planes (10 um); the Methods say only "the same z-depth".

while PG b , PG a , and 6G are spatially adjacent, they each reliably respond to unique combinations of odorants, with several pheromone/glomerulus combinations producing no detectable responses ... PG b did not respond to 6-methyl-5-hepten-2-one

Results, Alarm pheromone-responsive glomeruli are spatially stereotyped, paragraph 2; Figure 5C; Figure S6

With the disc, PGb responds at 48% to heptanone, heptanol and hexanol under both identifications, both plane readings and both models (plane a, model a: bix 0.851, 0.678, 0.734 with 5 ants; aiiy 0.596, 0.327, 0.528 with 12 ants), all with CIs above 0, and 6G responds to 6MHO (0.335 with 2 ants, 0.280 with 10). PGa is not identified under bix; under aiiy (4 ants) it responds to heptanol (0.258 to 0.274) and its hexanol prediction is 0.164 to 0.191 with the CI above 0, short of 0.2, so 5 of the 6 named combinations respond. PGb's prediction for 6MHO is negative with its CI below 0 (-0.087, -0.084), which the criterion counts as neither a response nor no response. The sets of PGb, PGa and 6G differ under the disc. With the region ROI, which takes in neighbours within 17.5 um, PGb responds to 6MHO in some fits and under aiiy the sets differ in one of four fits. No reading meets the whole criterion; the PGb and 6G parts hold. For the region ROI, 618 of the 1,684 traces had no disc pixel of at least 0.2 and fell back to the disc, which narrows the difference between the two ROIs.

How the test differed from the paper’s method

• The single-glomerulus traces are computed for two identifications, bix (the adopted M2 reading) and aiiy (the reading that identifies PGa in the most ants, chosen in the notebook by that rule), for both plane readings and both of the claims file's ROIs (the 15 um disc and the region of at least 0.2 at the centre). aiiy uses the ROI's maximum and drawing a, the reading set aside for the counts of figs5a-sparse-median-6, and 356 of drawing a's 436 regions are under 10 um across; it is used only because it finds PGa in the most ants, and the PGa results rest on it.
• The outcome rests on the disc ROI (the claims file's first M3 ROI and the typical glomerulus of Figure 5D), adopted as an assumption: the region of at least 0.2 is chosen on the same trial's peak map and absorbs neighbouring glomeruli.
• Each plane's row band is stabilized (phase correlation against the baseline mean, as for the maps); the unstabilized disc trace is kept for comparison (PGb peak fold change, median absolute difference 0.035, largest 0.591, r = 0.98; 3 of 129 trials change side of 0.2). Volume times include the centre plane's offset in the volume scan (0.0335 s per plane, from the files' starting_time).
• The model of reading (a) predicts for the average ant (sum-coded ant effects).

while responses in PG b and PG a peaked and then declined close to baseline by the end of the recording, responses in 6G were extremely slow, with a fluorescence plateau of tens of seconds

Results, Alarm pheromone-responsive glomeruli are spatially stereotyped, paragraph 2; Figure 5C

With the disc, under both identifications, both plane readings and both concentration readings, the mean PGb traces for heptanone, heptanol and hexanol end, over the last five volumes, at -0.165 to 0.116 with end over peak at most 0.151, and the mean 6G trace to 6MHO stays at or above half its peak for 36.1 to 37.3 s. PGa is identified only under aiiy (9 trials from 3 ants at 48%, 24 from 4 pooled); its heptanol and hexanol traces end at -0.149 to 0.046, but its hexanol trace peaks at only 0.129 to 0.148. With the region ROI the PGb heptanol trace ends at 0.272 to 0.472 and the criterion fails. The PGb and 6G parts hold; the PGa part rests on 3 or 4 ants under one identification, and with no clear hexanol response in PGa it passes trivially. For the region ROI, 618 of the 1,684 traces fell back to the disc (no disc pixel of at least 0.2).

How the test differed from the paper’s method

• The single-glomerulus traces are computed for two identifications, bix (the adopted M2 reading) and aiiy (the reading that identifies PGa in the most ants, chosen in the notebook by that rule), for both plane readings and both of the claims file's ROIs (the 15 um disc and the region of at least 0.2 at the centre). aiiy uses the ROI's maximum and drawing a, the reading set aside for the counts of figs5a-sparse-median-6, and 356 of drawing a's 436 regions are under 10 um across; it is used only because it finds PGa in the most ants, and the PGa results rest on it.
• The outcome rests on the disc ROI, as for fig5c-s6-tuning.
• Each plane's row band is stabilized (phase correlation against the baseline mean, as for the maps); the unstabilized disc trace is kept for comparison (PGb peak fold change, median absolute difference 0.035, largest 0.591, r = 0.98; 3 of 129 trials change side of 0.2). Volume times include the centre plane's offset in the volume scan (0.0335 s per plane, from the files' starting_time).

Sparse, unique subsets of AL glomeruli responded to all pheromones, while the paraffin oil vehicle did not generate responses

Results, Pheromone representation is sparse, paragraph 1; Figure 4B-C

Only biy meets the criterion. Under bix no vehicle ROI responds (largest vehicle ROI value 0.110), but heptanol and 6MHO have a responding ROI in only 6 of 13 ants each; under biy the pheromones give 12, 9, 13 and 11 of 13. Under the readings ii 7 to 10 ants have a vehicle-responding ROI (largest value 1.531); under aix and aiy one ant (ant2-m4-d26-y2022, 5 ROIs), whose vehicle trial has 13.5% of the AL at 0.2 or more against at most 2.4% in the other vehicle trials. No vehicle trial is flagged (more than half the AL at 0.2 or more), so leaving out the flagged trial gives the same vehicle counts.

How the test differed from the paper’s method

• The vehicle was given once per ant in the dataset, so its ROIs come from the pheromone trials.
• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.

We did not observe any fluorescence decreases in response to odor, although we did detect small, non-specific decreases in fluorescence due to minor shifts in AL position and photobleaching

Results, Pheromone representation is sparse, paragraph 1

594 regions at least 10 um across reach -0.2 or below in at least two trials of one odor at the higher concentration, in the same plane, without doing so in the vehicle trial (540 of them 10 to 20 um), in all 13 ants; the most negative has a mean of -0.452 (ant4-m4-d26-y2022, hexanol, plane 3). 168 more are negative in the vehicle trial too. Their placement points to motion, which the paper names as the source of small non-specific decreases: 487 of the 594 touch the AL border (within one 1.2 um bin) and 564 lie within 5 um of it, against 83% expected for a 10 um region placed at random in the AL; 384 overlap a negative region of another odor in the same plane (302 by half or more). The criterion counts them all, so the outcome follows the reader's rule; the vehicle comparison rests on one trial per ant.

How the test differed from the paper’s method

• Only each ant's higher concentration was mapped for this claim.
• Stabilization by whole-pixel phase correlation, not Image Stabilizer; residual motion could produce some of the negative regions, and the vehicle comparison removes only those it shares.
• A region "doing so in the vehicle trial" was read as at least half of its bins at -0.2 or below in the vehicle map of the same plane; the claims file does not say how a region is matched.

Fluorescence increases were frequently large (1–2-fold change) and lasted longer than the 5s odor presentation

Results, Pheromone representation is sparse, paragraph 1

No reading meets both parts. With glomerulus-sized ROIs (drawing b) the median peak fold change of responding ROI-trials is 0.453 with the mean value (445 ROI-trials, 8.8% within 1 to 2, 70.3% still above baseline mean plus 3 SD at 8.4 s) and 1.016 with the maximum (1,086, 32.7% within 1 to 2, 41.3% lasting). With drawing a the medians are 0.377 and 0.573. The criterion needs both value readings in 0.9 to 2.2 and more than half lasting.

How the test differed from the paper’s method

• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.
• The trace for "lasting" is the ROI's mean raw fluorescence at its depth plane, on 4 x 4 binned volumes 1 to 5 and 8, stabilized as the maps are.

we observed consistent partial overlap in the response patterns activated by the three compounds eliciting panic alarm responses ... with a single glomerulus activated by all three

Results, Pheromone representation is sparse, paragraph 2; Figure S5B

Only biy at 48% meets the criterion (median 1, exactly one shared ROI in 5 of 8 ants). Under bix the median number of ROIs responding to heptanone, heptanol and hexanol at one depth is 0 at the higher concentration (exactly one in 4 of 13 ants, none in 8) and 1 at 48% (exactly one in 4 of 8). aix and aiy give a median of 1 with exactly one in 5 or 6 of 13 ants; the readings ii medians of 2 to 2.5. Heptanone and hexanol share an ROI in 9 of 13 ants under bix. Leaving out the flagged trial gives the same.

How the test differed from the paper’s method

• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.

while we sometimes observed responses to 6-methyl-5-hepten-2-one and either 4-methyl-3-heptanone or 4-methyl-3-hexanol in an overlapping region, those occurrences were rare and inconsistent

Results, Pheromone representation is sparse, paragraph 2; Figure S5C

The criterion asks for the result under each M2 reading, and the four readings with the ROI maximum (aiix, aiiy, biix, biiy) fail at both concentration readings, with medians 1 to 3 and shared ROIs in 11 of 13 ants at the higher concentration (met under 4 of 8 readings, under every reading False). The four readings with the ROI mean (aix, aiy, bix, biy) meet it; under bix the median number of ROIs shared by 6MHO and another pheromone is 0, with any in 1 of 13 ants (1 of 8 at 48%). With heptanone or hexanol only (the statement's wording), and without the flagged trial, the result is the same. The outcome was reproduced, resting on the ROI mean, until the third verification round agreed to the change. Under drawing a the maximum-value readings call a response from one pixel of at least 0.2 in mostly sub-glomerular regions (ant1-m4-d27 has 27 shared ROIs under biix and none under bix), so the failure may come from that reading as much as from the data.

How the test differed from the paper’s method

• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.

The three pheromones with overlapping calcium response patterns all robustly elicited panic alarm behavior, while 6-methyl-5-hepten-2-one did not elicit panic alarm behavior and generated a non-overlapping response

Results, Alarm pheromone-responsive glomeruli are spatially stereotyped, paragraph 4; Figure 6

Imaging half only. figs5b-shared-glomerulus is not met under any reading at the higher concentration and only under biy at 48%; figs5c-mho-rarely-overlaps is met under the four mean-value readings and fails under the four maximum-value readings, so it is not reproduced. The behaviour half rests on fig1e-categorical-response, whose data are not in the dataset.

How the test differed from the paper’s method

• Judged on figs5b-shared-glomerulus and figs5c-mho-rarely-overlaps as tested here.

PG b is located approximately halfway between the dorsal and ventral AL surfaces, and is neighbored by two additional glomeruli that respond to alarm pheromones, with all three visible in the same optical plane ... Both glomeruli were identified in 13/13 individuals. In 11/13 individuals, we identified a third neighboring glomerulus

Results, Alarm pheromone-responsive glomeruli are spatially stereotyped, paragraph 1; Figure 5A-B

No M2 reading gives all three counts within tolerance; PGb is identified in 5 to 12 of 13 ants, 6G in 1 to 10 and PGa in 0 to 4 (bix 5, 2, 0, given as measured; aiix 12, 10, 3; aiiy 12, 10, 4). With the deepest AL plane where the AL's baseline fluorescence falls to 25% of its brightest plane, PGb lies in the middle third of the depth in 1 to 3 ants per reading (bix 1 of 5; the other 4 at 0.71 to 0.77 of the depth). With a 10% rule 5 to 11 ants per reading are in the middle third.

How the test differed from the paper’s method

• The identification follows M4 with "adjacent" read as centres at most 20 um apart (two touching glomeruli of at most 20 um, Results); the claims file does not give a distance.
• The deepest plane with AL fluorescence was read with a 25% rule, with 10% and 50% shown; the claims file does not give one.
• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.

PG a was always located anterior (mean distance between centers: 12.9±1.9SD μm), and slightly lateral (mean distance: 5.1±2.9SD μm) to PG b ( Fig. 5D ). In comparison, 6G was always lateral to PG b (mean distance: 13.1±2.6SD μm), and in a similar position along the anterior-posterior axis (mean distance: 0.6±2.2SD μm)

Results, Alarm pheromone-responsive glomeruli are spatially stereotyped, paragraph 3; Figure 5D

No reading meets the criterion, so the readings do not disagree on the outcome. Under bix no PGa is identified, and 6G (2 ants) lies 12.9 to 15.6 um lateral and 0.1 to 3.5 um anterior of PGb on average under the centre readings ac, ad, bc and bd. Under the other readings PGa (2 to 4 ants) lies 9.2 to 18.5 um anterior on average (expected 12.9 +/- 1.9) with SDs up to 13.1 um, and 0.4 to 9.8 um lateral with SDs up to 14.0 um; 6G (1 to 10 ants) lies 7.6 to 20.8 um lateral with SDs of 1.3 to 8.8 um. The image axes are those the files give ("up is lateral, right is anterior"); a baseline image of ant1-m4-d26-y2022 with the three centres marked shows PGa anterior and 6G lateral of PGb, as in Figure 5B. The direction conditions do not hold in every ant under every reading, since the centres are taken again after identification; PGa is not anterior in every ant under aiiy bc, biix bd and biiy bd, and 6G is not lateral in every ant under aiix bd, aiiy ad and bd, aiy ad and bd, biix ac and biiy ac.

How the test differed from the paper’s method

• Centres were taken on 4 x 4 binned per-plane maps (1.16 um) in PGb's plane.
• M4 identifies PGa and 6G by their direction from PGb, but the centres measured here are taken again on the defining odors' maps, so the direction conditions are tested and do not hold by construction (round 1 said they did, which the outputs do not show).

Fluorescence plateaued for ~30s before declining and returning to baseline ~80s after odor presentation

Results, Alarm pheromone-responsive glomeruli are spatially stereotyped, paragraph 2; Figure S7A

Under aiiy 6G is identified in all 5 ants with a 144 s recording (under bix in 1). With the disc the mean trace peaks at 0.252 to 0.281 at 33.1 s; it stays at or above 90% of its peak for 2.4 to 3.6 s and 80% for 4.8 to 8.4 s, against about 30 s, and returns 27.5 to 56.6 s after odor presentation, against 80 +/- 8 s (measured gives aiiy, disc, plane a, reading ade: the 80% plateau and the return below baseline mean plus 3 SD from odor onset). With the region ROI the return is 70.8 to 91.5 s but the plateau 7.2 to 15.7 s. Under the disc one trial returns more than 100 s after odor presentation (100.7 to 105.7 s by onset reading); under the region ROI with return reading b two trials do (105.9 to 110.9 s and 100.7 to 105.7 s). For the region ROI, 618 of the 1,684 traces fell back to the disc (no disc pixel of at least 0.2). None of the 64 combinations meets the criterion.

How the test differed from the paper’s method

• The single-glomerulus traces are computed for two identifications, bix (the adopted M2 reading) and aiiy (the reading that identifies PGa in the most ants, chosen in the notebook by that rule), for both plane readings and both of the claims file's ROIs (the 15 um disc and the region of at least 0.2 at the centre). aiiy uses the ROI's maximum and drawing a, the reading set aside for the counts of figs5a-sparse-median-6, and 356 of drawing a's 436 regions are under 10 um across; it is used only because it finds PGa in the most ants, and the PGa results rest on it.
• Each plane's row band is stabilized (phase correlation against the baseline mean, as for the maps); the unstabilized disc trace is kept for comparison (PGb peak fold change, median absolute difference 0.035, largest 0.591, r = 0.98; 3 of 129 trials change side of 0.2). Volume times include the centre plane's offset in the volume scan (0.0335 s per plane, from the files' starting_time).

Time to response onset: significant effects of pheromone (p=0.0034), glomerulus (p<0.0001), and the interaction (p=0.0030). Time to response maximum: significant effects of pheromone (p<0.0001) and glomerulus (p=0.047), but not the interaction (p=0.88)

Results, Alarm pheromone-responsive glomeruli are spatially stereotyped, paragraph 2; Figure S7C legend

What the paper reports

• onset, pheromone: linear mixed effects model (lme), p = 0.0034 (n = 8 ants, three trials per condition per ant, 48%)
• onset, glomerulus: linear mixed effects model (lme), p < 10⁻⁴ (n = 8 ants, three trials per condition per ant, 48%)
• onset, pheromone by glomerulus interaction: linear mixed effects model (lme), p = 0.003 (n = 8 ants, three trials per condition per ant, 48%)
• time to maximum, pheromone: linear mixed effects model (lme), p < 10⁻⁴ (n = 8 ants, three trials per condition per ant, 48%)
• time to maximum, glomerulus: linear mixed effects model (lme), p = 0.047 (n = 8 ants, three trials per condition per ant, 48%)
• time to maximum, interaction: linear mixed effects model (lme), p = 0.88 (n = 8 ants, three trials per condition per ant, 48%)

Under bix no PGa is identified, so the models cannot be fitted (one glomerulus level; the notebook lists each as NOT TESTED). Under aiiy, which identifies PGa in 4 of 13 ants (3 at 48%), the models fit with 18 PGa and 48 PGb traces at 48% (with the disc 8 or 9 PGa traces never exceed the bound and are excluded, with the region 2 or 3; 1 PGb trace with the disc). Across the 12 combinations of ROI, plane and p-value reading, 1 to 4 of the 5 effects have p < 0.05, and the time-to-maximum interaction has p < 0.05 in 9 of 12, where the paper reports p = 0.88. Of the 24 sets of fits, 23 converged; the ML fits of reading (a) for onset with the region ROI and plane a did not. The evidence is weak, with PGa from 3 ants against the paper's 8.

How the test differed from the paper’s method

• The single-glomerulus traces are computed for two identifications, bix (the adopted M2 reading) and aiiy (the reading that identifies PGa in the most ants, chosen in the notebook by that rule), for both plane readings and both of the claims file's ROIs (the 15 um disc and the region of at least 0.2 at the centre). aiiy uses the ROI's maximum and drawing a, the reading set aside for the counts of figs5a-sparse-median-6, and 356 of drawing a's 436 regions are under 10 um across; it is used only because it finds PGa in the most ants, and the PGa results rest on it.
• The outcome rests on the aiiy identification, the only one under which the claim can be tested; under the adopted bix it cannot.
• Each plane's row band is stabilized (phase correlation against the baseline mean, as for the maps); the unstabilized disc trace is kept for comparison (PGb peak fold change, median absolute difference 0.035, largest 0.591, r = 0.98; 3 of 129 trials change side of 0.2). Volume times include the centre plane's offset in the volume scan (0.0335 s per plane, from the files' starting_time).

Higher concentrations produced more robust responses, with a few more glomeruli passing the threshold, but overall spatial response patterns were similar across concentrations

Results, Pheromone representation is sparse, paragraph 2; Figure S5A; Figure S6

Under bix the count at the higher concentration exceeds the lower in 28 of 52 ant and pheromone pairs, falls short in 2, ties in 22 (two-sided sign test p < 0.0001), and every ROI responding at the lower concentration responds at the higher in 50 of 52 pairs. aix, aiy, biy and biiy also meet the criterion; aiix and aiiy pass the sign test but fail the subset condition (21 and 23 of 52), and biix fails the sign test (14 against 12, p = 0.845). The ROIs are drawn at the higher concentration, as in the Methods, which favours ROIs that respond there.

How the test differed from the paper’s method

• The outcome rests on the adopted M2 reading (ROI mean, trial-mean counting; see figs5a-sparse-median-6), under which both drawings meet the criterion.
• ROIs are the regions of at least 0.2 in a trial's map that recur in another trial of the odor (claims file, drawing a), a recurrence read as an overlap of at least half the smaller region, an assumption: the claims file does not say how much overlap makes a recurrence. In round 1 the pixels of at least 0.2 in two trials were taken instead, which broke the maps into fragments.