CRCNS:ret-1 · Encore Science

Claims

From the paper 10.1242/dev.027912

Reproducing Lefebvre et al. (2008), Figure 9, from CRCNS ret-1 7

Notebook reproduce_lefebvre_2008.ipynb

Not reproducedFigure 9 legend, panels C to H
Figure 9C to 9H report 115 RGCs from mutant retinas and 143 RGCs from control retinas.
plots a cumulative histogram of that quantity for RGCs in mutant retinas (red curve, 115 cells ... and control retinas (black curve, 143 cells
Figure 9 legend, panels C to H
The data set's labels give 200 mutant units (87 Chx10-Cre in 6 retinas, 113 Pax6-Cre in 4) and 122 control units (95 labelled "WT (chx10 het)" in 3 retinas, 27 "WT (pax6 het)" in 2), against 115 (103.5 to 126.5) and 143 (128.7 to 157.3). Every unit fired during the bar stimulus. Units whose STA peak exceeds 1.5 times the null's number 138 mutant and 113 control; above 2 times, 103 and 107. Neither mutant count shown is inside its tolerance (103 is just below 103.5), but the mutant count falls from 138 to 103 between the two thresholds, so a threshold between them could put it inside, and with the 20080702 frames as deposited the count above 1.5 is 126, inside it. No count of control units can exceed 122, so the counts miss under any reading of a flicker response because of the control count. Without the second unit of each duplicate pair, 189 and 118. The Bax -/- retina (3 units) is in neither group. The notebook tabulates the units per retina and label; the files do not give the region recorded in the Pax6-Cre retinas, so the paper's peripheral subset cannot be formed.
How the test differed from the paper’s method
- The claims file asks also for the count of units with a flicker response without defining one; the counts above two thresholds of the STA peak over a null STA are shown as a standard choice, and none can change the outcome.
- The control retinas are as the data set labels them; 95 of the 122 control units carry a label ("WT (chx10 het)") that the paper's description of its controls does not include.
Not reproducedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9G
In a Linear-Nonlinear model fitted to the flicker responses, the response threshold of mutant RGCs is unaltered.
Based on fitting with a Linear-Nonlinear model ... we found that the threshold is unaltered, but the gain is reduced in mutants ( Figures 9G-H )
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9G
In all 36 combinations (3 recording readings, 2 fits, 2 forms, 3 bands) the bands separate, with mutant thresholds higher, DKW included; reading A, frame fit, median 1.12 against 0.44 SD of the generator (continuous form) and 1.69 against 0.01 (printed form). Part of this comes from mutant units with weak STAs, whose generator predicts little of the rate; among units with clear STAs the medians are 0.86 against 0.41 (reading A, frame fit, continuous form), and over the three readings 24 of 36 combinations separate and no DKW one, which the criterion would call blocked. The outcome depends on using all units, as the claims file's header reads the groups. With the frames as deposited and with 16.6 um bars alone all 36 separate. The binned fit in the printed form stops at a grid edge in 107 of 200 mutant fits, all at the top, and 39 of 122 control fits, 33 at the top and 6 at the bottom (reading A); why they reach the top is not shown. Without the edge fits its medians are -0.03 and -0.03 (KS p = 0.21) and its DKW and bootstrap bands overlap, but that comparison drops most high mutant fits by construction and is no evidence that the threshold is unaltered. The frame fit in the continuous form has 1 fit per group at a grid edge, at the top, and separates under every band. With the Pax6-Cre heterozygous controls alone the DKW bands still separate.
How the test differed from the paper’s method
- Theta is searched on 99 quantiles of y for the frame fit and at every split between the 20 bin centres for the binned fit, a standard choice of resolution; the binned grid cannot place theta above the last split, which many printed-form fits reach.
- The vertical bars of 20080702_R1 and 20080702_R4 are rebuilt with a frame of 16.6718 ms, as for fig9d-rf-size-similar.
ReproducedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3; Supplementary Figure 3 (not read)
Some control and mutant RGCs have receptive field surrounds of opposite sign to the centre, so lateral inhibition functions in mutant circuits.
Some control and mutant RGCs had receptive field surrounds, where light has the opposite effect of the center
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3; Supplementary Figure 3 (not read)
Under all 12 readings (3 recording readings, 2 separations, depths of 10% and 20%) at least one mutant and one control unit have a lobe of opposite sign on both flanks of b(x); reading A, svd, 20%, 90 of 200 mutant and 16 of 122 control units. The same rule applied to null STAs (spike counts shifted by half the recording) is met by 194 to 198 of 200 mutant and 119 to 121 of 122 control units, so the rule is met by noise. Among units with an STA peak above twice the null's, at least 7 mutant and 4 control units meet it at 20% in every reading. The outcome follows the criterion, but the result is no evidence for or against surrounds.
How the test differed from the paper’s method
- The vertical bars of 20080702_R1 and 20080702_R4 are rebuilt with a frame of 16.6718 ms, as for fig9d-rf-size-similar.
ReproducedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9E
The time course of the light response is significantly slower in mutant retina, measured as the time to peak of the temporal integration function a(t).
On the other hand, the time course of the light response was significantly slower in mutant retina ( Figure 9E )
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9E
In all 36 combinations (3 recording readings, 2 separations, 2 peak readings, 3 bands) the bands separate with the mutant curve at longer times, and in none do they separate the other way; mutant medians exceed control ones by 16.7 to 33.3 ms (reading A, svd, frame, 125 against 108 ms), KS p at most 6e-10. The DKW band, which never separated under shuffled labels, separates in every combination. The result holds in every variant, including units with clear STAs (103 mutant, 107 control), each Cre line alone and the Pax6-Cre heterozygous controls alone under reading A, and among units with clear STAs, with 16.6 um bars alone and with the 20080702 frames as deposited under all three readings (36 of 36 separate in each). With the frames as deposited, 8 binomial and bootstrap combinations (peak separation, readings A and C) also hold 1 or 2 values where the bands separate the other way; with the frames corrected none do. The statement's "significantly" has no named test in the paper; the KS test is shown.
How the test differed from the paper’s method
- Lag j is placed at (j + 0.5) frames, the data set's example script's convention; the offset is the same for every unit.
- The vertical bars of 20080702_R1 and 20080702_R4 are rebuilt with a frame of 16.6718 ms, as for fig9d-rf-size-similar.
ReproducedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9F
Mutant RGCs fire at much lower rates than control RGCs in response to flickering gratings.
Moreover, mutant RGCs fired at much lower rates in reponse to flicker stimuli ( Figure 9F )
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9F
In all 9 combinations (3 recording readings, 3 bands) the bands separate with the mutant curve at lower rates, DKW included; reading A, median 3.82 against 9.02 Hz (ratio 0.42; 0.38 to 0.45 over readings). By retina, mutant medians 1.9 to 15.9 Hz, control 3.8 to 23.6 Hz. Every variant gives a lower mutant median; with Pax6-Cre mutants alone (ratio 0.68) and among units with clear STAs (0.66), the DKW bands overlap while the binomial and bootstrap bands separate (those two separate in 38 and 39% of label shuffles for this measure). With the Pax6-Cre heterozygous controls alone (ratio 0.41) the DKW bands separate. The outcome depends on using all units, as the claims file's header reads the groups (all units of each group, the paper naming no exclusion). Among units with clear STAs, over the three readings, 6 of 9 combinations separate and no DKW one, which the criterion would call blocked. With the frames as deposited and with 16.6 um bars alone all 9 separate. The answer to the question in questions/crcns-ret-1.yaml (which bar recording) found that the paper does not say; the criterion requires the three recording readings A, B and C, with the pooled spike count and duration under B, as run here.
BlockedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9D· Methods underspecified
Receptive field centre size varies greatly among RGCs, but its distribution is similar in control and mutant retinas.
The size of the receptive field center varied greatly among RGCs, but the distribution was similar in control and mutant retinas ( Figure 9D )
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9D
The outcome rests on one borderline combination and on the frame correction of the 20080702 bar recordings. The readings disagree. In 17 of 18 combinations (3 recording readings, 2 separations, 3 bands) the 95% bands separate with mutant centres smaller; under reading B with the svd separation the DKW bands overlap (median 134 against 156 um, KS p = 0.004). Reading A, svd, median 122 against 156 um (KS p below 1e-4). This follows the correction of the frame of the 20080702 bar recordings; with the frames as deposited the reading B DKW bands separate too and all 18 separate, which was the outcome (not_reproduced) of the first run. Much of the difference comes from mutant units with weak STAs (97 of 200 mutant and 15 of 122 control units have an STA peak less than twice the null's). Among units above that, the medians are 145 and 158 um (reading A, KS p = 0.19), the binomial and bootstrap bands separate only between 82 and 101 um, and the DKW bands overlap; over the three readings 13 of 18 combinations separate in that subset, and 1 of the 6 DKW ones, so the outcome there is blocked too. With the 16.6 um bars alone 16 of 18 separate (the svd DKW bands overlap under readings A and B), and with Pax6-Cre mutants alone the svd DKW bands overlap under reading A. Under shuffled group labels the binomial and bootstrap bands of this measure separate in 38 and 45% of shuffles and the DKW band in none. The answer to the question in questions/crcns-ret-1.yaml (which bar recording) found that the paper does not say, and the criterion requires the outcome under the three recording readings A, B and C, which are run here.
How the test differed from the paper’s method
- All 5 retinas recorded with 8.3 um bars are mutant; every control retina was recorded with 16.6 um bars.
- The mutant group cannot be restricted to the peripheral region of the Pax6-Cre retinas, as the data set does not say where they were recorded.
- The vertical bars of 20080702_R1 and 20080702_R4 are rebuilt with a frame of 16.6718 ms, the frame of the other bar recordings, not the stated 16.6691 and 16.6653 ms (see findings).
BlockedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9H· Methods underspecified
In the same Linear-Nonlinear model, the gain of mutant RGCs is reduced.
we found that the threshold is unaltered, but the gain is reduced in mutants ( Figures 9G-H )
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 4; Figure 9H
The readings of the nonlinearity's form disagree. With the printed form (B + G y above theta), mutant gains are lower and the bands separate in all 18 combinations (reading A, frame fit, median 18.4 against 23.1 Hz per SD, ratio 0.80). With the continuous form (B + G (y - theta)), the bands overlap in 9 of 18, separate with mutant gains lower in 4 (reading B, binomial and bootstrap) and the other way in 5 (readings A and C); the DKW bands overlap in all 6 continuous-form combinations (reading A, frame fit, median 30.4 against 30.8, KS p = 0.76). Under reading B the continuous-form separations rest mostly on the 4 to 13 mutant gains below the lowest control gain, and under reading A, frame fit, bootstrap, on 1 mutant against 6 control cells; the others lie inside the distributions (reading A, binned fit, 165 of 200 mutant and 113 of 122 control cells at or below 41.46; reading C, frame fit, 16 to 19 mutant against 23 to 25 control cells near 21.2). Among units with clear STAs neither form gives lower mutant gains under reading A (printed 23.1 against 23.4, continuous 33.1 against 31.0); over the three readings in that subset the printed form separates with mutant gains lower in 6 of 18 combinations and the continuous form in 4, with 6 the other way and no DKW separation, so the outcome there is blocked too. With the frames as deposited the continuous form separates in 4 of 18, all with mutant gains lower, and the outcome is the same. On synthetic cells that follow the continuous form (theta 0.5, G 20), the printed-form frame fit returns theta 0.75 and G 13.4, so part of the printed-form gain reduction may mirror the higher mutant thresholds.
How the test differed from the paper’s method
- The vertical bars of 20080702_R1 and 20080702_R4 are rebuilt with a frame of 16.6718 ms, as for fig9d-rf-size-similar.

Not tested in a notebook 19

19 data missing from the dataset

BlockedResults, Broad expression of gamma protocadherins in retina, paragraphs 2 and 3; Figure 1· Data missing from the dataset
Pcdh-gamma proteins (Pcdh-gamma-GFP in fusg mice) are present in all five retinal layers and in all six retinal cell types (photoreceptors, horizontal, bipolar, amacrine, RGCs and Mueller glia).
As judged by localization of GFP in Pcdh-γ fusg mice, Pcdh-γs are present in all five retinal layers ... All six cell types were Pcdh-γ-positive ... Thus, Pcdh-γs are expressed in all cell types of the neural retina.
Results, Broad expression of gamma protocadherins in retina, paragraphs 2 and 3; Figure 1
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Broad expression of gamma protocadherins in retina, paragraph 4; Figure 2· Data missing from the dataset
Pcdh-gamma is present in the IPL, neuroblast layer and RGC axons at P0, appears in the developing OPL by P7, and reaches the adult pattern by P14.
At this time, Pcdh-γ is present on cells in the neuroblast layer, in the IPL, and on RGC axons ( Figure 2A ) ... By P7, Pcdh-γ appears in the OPL, as it divides the neuroblast layer into INL and ONL ( Figure 2C ). By P14, the adult pattern described above is established
Results, Broad expression of gamma protocadherins in retina, paragraph 4; Figure 2
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Synaptic localization of Pcdh-γ proteins in the retina, paragraphs 1 and 2; Figure 3· Data missing from the dataset
Pcdh-gammas are present in pre- and postsynaptic compartments of rod and cone synapses in the OPL, though little is in horizontal cell varicosities, and at many but not only synaptic sites in the IPL.
Thus, Pcdh-γs were present in pre- and post-synaptic compartments of rod and cone synapses. In contrast, although horizontal cell processes labeled with anti-calbindin ... were Pcdh-γ-positive, little Pcdh-γ was present in their synaptic varicosities
Results, Synaptic localization of Pcdh-γ proteins in the retina, paragraphs 1 and 2; Figure 3
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Inactivation of Pcdh-γs in the retina leads to neuronal and synaptic loss, paragraph 4; Supplementary Figure 1A and 1B (not read)· Data missing from the dataset
Chx10-Cre recombination is extensive (above 90%) in the INL and ONL, while about half of the cells in the GCL are not recombined.
Recombination was extensive (>90%) in the INL and ONL but occasional columns of cells were spared ... In contrast, approximately half of the cells in the GCL were GFP-negative and LacZ-positive
Results, Inactivation of Pcdh-γs in the retina leads to neuronal and synaptic loss, paragraph 4; Supplementary Figure 1A and 1B (not read)
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Inactivation of Pcdh-γs in the retina leads to neuronal and synaptic loss, paragraph 5; Figure 4B to 4H· Data missing from the dataset
At P18, retinas lacking Pcdh-gammas in Chx10-Cre mutants are properly laminated but about 25% thinner than controls, from a selective reduction of about 40% in INL and IPL thickness.
However, mutant retinas were ∼25% thinner than those of wild-type mice or heterozygote littermates. The difference resulted from a selective reduction of ∼40% in the thickness of the INL and the IPL ( Figure 4H ).
Results, Inactivation of Pcdh-γs in the retina leads to neuronal and synaptic loss, paragraph 5; Figure 4B to 4H
What the paper reports
- ANOVA and post-hoc Tukey test, p < 0.01 (n = 3-4 animals of each genotype)
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Inactivation of Pcdh-γs in the retina leads to neuronal and synaptic loss, paragraph 5; Figure 4D to 4H· Data missing from the dataset
The INL and IPL are thinned to the same extent in Chx10-Cre;Pcdh-gamma fcon3/fcon3 and fdel/fdel retinas, so fcon3 is functionally a null.
The INL and IPL were thinned to the same extent in Chx10-Cre;Pcdh-γ fcon3/fcon3 and Chx10-Cre;Pcdh-γ fdel/fdel mice
Results, Inactivation of Pcdh-γs in the retina leads to neuronal and synaptic loss, paragraph 5; Figure 4D to 4H
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Inactivation of Pcdh-γs in the retina leads to neuronal and synaptic loss, paragraph 6; Figure 4I· Data missing from the dataset
In peripheral Pax6alpha-Cre;Pcdh-gamma fcon3/fcon3 retina at P18, bipolar, amacrine and retinal ganglion cells are reduced by 45 to 65%, Mueller glia by about 20%, while horizontal cells and photoreceptors differ little from controls.
Numbers of bipolar, amacrine, and retinal ganglion cells were reduced by 45-65% ( Figure 4I ). Müller glia were also decreased, but only by ∼20%. In contrast, numbers of horizontal cells and photoreceptors differed little between mutants and controls
Results, Inactivation of Pcdh-γs in the retina leads to neuronal and synaptic loss, paragraph 6; Figure 4I
What the paper reports
- Mann-Whitney non-parametric test, p < 10⁻⁴ (n = 6-8 animals of each genotype)
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Increased postnatal apoptosis in the absence of Pcdh-γs, paragraph 1; Figure 5A to 5I· Data missing from the dataset
Mutant and control retinas do not differ perinatally (E17.5 to P3); INL and IPL are thinner in mutants by P7, about 40% thinner by P14, and change little over the following months.
We detected no differences in laminar arrangement or thickness between mutant and control retinas perinatally ... Both layers were ∼40% thinner in mutants than in controls by P14, then changed little over the following several months
Results, Increased postnatal apoptosis in the absence of Pcdh-γs, paragraph 1; Figure 5A to 5I
What the paper reports
- Student's t-test, p < 0.001 (n = 3-4 animals per stage)
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Increased postnatal apoptosis in the absence of Pcdh-γs, paragraph 2; Figure 5J to 5N· Data missing from the dataset
Apoptosis (cleaved caspase-3) follows the normal time course but is significantly higher in mutant than control retinas during the first postnatal weeks.
Apoptosis followed a similar time course in Pcdh-γ deficient retinas, but levels were significantly higher in mutants than in controls ... Although increased apoptosis was seen in both neuroblast and ganglion cell layers at P0, it was confined to the INL at P7
Results, Increased postnatal apoptosis in the absence of Pcdh-γs, paragraph 2; Figure 5J to 5N
What the paper reports
- Student's t-test or Mann-Whitney test, p < 0.05 (n = 3-6 animals per stage)
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Cell autonomy of Pcdh- γ -dependent cell survival in retina, paragraph 2; Figure 6B, 6C and 6F· Data missing from the dataset
In central retina of Pax6alpha-Cre;Pcdh-gamma fcon3/fcon3 mice, where only the Pax6alpha amacrines lack Pcdh-gammas, these amacrines are 58% fewer than in controls, a loss equivalent to that in peripheral retina.
The number of Pax6α-amacrines in central retina was 58% lower in mutants than in controls ... loss of Pax6α-positive amacrines is equivalent in peripheral and central retina
Results, Cell autonomy of Pcdh- γ -dependent cell survival in retina, paragraph 2; Figure 6B, 6C and 6F
What the paper reports
- Pax6alpha amacrines, central retina, mutant against control; the weakest mark the legend defines, as the text does not say which mark this bar carries: Student's t-test or Mann-Whitney test, p < 0.05 (n = 6 to 8 animals per genotype)
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Cell autonomy of Pcdh- γ -dependent cell survival in retina, paragraph 3; Figure 6F· Data missing from the dataset
Loss of Pcdh-gammas from Pax6alpha amacrines in central retina does not affect the survival of bipolar cells, horizontal cells, Mueller glia or starburst amacrines, but causes a small (about 15%) significant loss of Brn3a-positive RGCs.
Loss of Pcdh-γs from the Pax6α-positive amacrines had no detectable effect on the survival of bipolar cells, horizontal cells or Müller glia ... In contrast, we detected a small (∼15%) but significant loss of Brn3a-positive RGCs from central retina
Results, Cell autonomy of Pcdh- γ -dependent cell survival in retina, paragraph 3; Figure 6F
What the paper reports
- Brn3a+ RGCs, central retina, mutant against control: Student's t-test or Mann-Whitney test, p < 0.05 (n = 6 to 8 animals per genotype)
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Pcdh- γ s are dispensable for laminar targeting of retinal neurons, paragraph 1; Figure 7· Data missing from the dataset
In Pcdh-gamma mutant retina at P18, processes of all ten examined amacrine and bipolar subtypes are arrayed in their appropriate IPL sublaminae, though with occasional gaps.
In all ten cases, processes were arrayed in appropriate sublaminae in mutant retinas at P18, although disruptions or gaps were sometimes present
Results, Pcdh- γ s are dispensable for laminar targeting of retinal neurons, paragraph 1; Figure 7
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Synapses and arbors of Pcdh-γ-deficient neurons destined to die, paragraph 1; Figure 8A to 8D· Data missing from the dataset
Deleting Bax rescues the neurons that die in Pcdh-gamma mutant retina; INL and GCL thickness and Chx10+ bipolar number do not differ between Pcdh-gamma-positive and -negative Bax-/- retinas, and the IPL expands to the Bax-/- thickness.
Likewise, Bax deletion rescued neurons destined to die in Pcdh-γ mutant retina: the thickness of the INL and GCL and the number of Chx10-positive bipolar cells were indistinguishable
Results, Synapses and arbors of Pcdh-γ-deficient neurons destined to die, paragraph 1; Figure 8A to 8D
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Synapses and arbors of Pcdh-γ-deficient neurons destined to die, paragraph 2; Figure 8E to 8L; Supplementary Figure 2 (not read)· Data missing from the dataset
Synapse density in the IPL and OPL (PNA or Bassoon staining) does not differ significantly between Pcdh-gamma-positive and Pcdh-gamma-negative Bax-/- retinas.
The density of synapses in the IPL and OPL, as judged by staining for PNA or Bassoon, did not differ significantly between Pcdh-γ-positive and Pcdh-γ-negative Bax −/− retinas
Results, Synapses and arbors of Pcdh-γ-deficient neurons destined to die, paragraph 2; Figure 8E to 8L; Supplementary Figure 2 (not read)
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Synapses and arbors of Pcdh-γ-deficient neurons destined to die, paragraph 3; Figure 8M and 8N· Data missing from the dataset
Neurons rescued from death by Bax deletion target their processes to the appropriate IPL sublaminae as specifically as in Pcdh-gamma single mutants, and the gaps of the single mutants are absent.
In all cases, targeting of processes to appropriate laminae was as specific in double mutants as in Pcdh-γ single mutants
Results, Synapses and arbors of Pcdh-γ-deficient neurons destined to die, paragraph 3; Figure 8M and 8N
The claim rests on histology (sections, dissociated cells or counts of stained cells). The data set holds only spike times and white-noise stimuli from the multielectrode recordings, and the paper names no other deposit of its images or counts.
BlockedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3; Figure 9A· Data missing from the dataset
RGCs in mutant and control retinas show a similar variety of responses to small flashing spots, including sustained and transient ON, OFF and ON-OFF responses.
RGCs in Pcdh-γ mutant and control retinas showed a similar variety of responses to small flashing spots, including sustained and transient ON, OFF and ON-OFF responses
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3; Figure 9A
The data set holds no flashing-spot recordings; every recording is binary white noise (flickering bars or checkerboards). The claim also depends on fig9c-on-off-index, blocked for the same reason. The paper names no other deposit.
BlockedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3; Figure 9C· Data missing from the dataset
The proportions of ON- and OFF-dominated responses are the same in mutant and control retinas.
Proportions of ON- and OFF-dominated responses were identical in mutants and controls ( Figure 9C )
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3; Figure 9C
The On-Off index needs the flashing-spot experiment of Figure 9A, which is not in the data set; every recording is binary white noise.
BlockedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3 (data not shown)· Data missing from the dataset
Mutant RGCs respond to very dim flashes, which excite only rods, and to bright flashes, which mainly excite cones, so both rod and cone pathways function.
Mutant RGCs responded to very dim flashes, which excite only rods, and also to bright flashes, which predominantly excite cones (data not shown)
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3 (data not shown)
The data set holds no flashes, dim or bright; every recording is binary white noise at one luminance range, so the criterion's pair of intensities is absent.
BlockedResults, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3; Figure 9B· Data missing from the dataset
Both ON and ON-OFF direction-selective RGCs are found in mutant retinas.
Both ON and ON-OFF direction-selective cells were encountered in mutant retinas ( Figure 9B and data not shown)
Results, Functional visual circuits form in the absence of Pcdh- γ s, paragraph 3; Figure 9B
The data set holds no moving gratings or moving bars; every recording is binary white noise. It also depends on fig9a-response-variety, which is blocked.

Findings about the data

What attempts found about the dataset itself, for its owner and the archive.

Data missingThe data set holds only the white-noise recordings behind Figure 9D to 9H. The histology of Figures 1 to 8 and Supplementary Figure 1, and the flashing-spot, flash, moving-bar and grating recordings of Figure 9A to 9C, are not in it, and the paper names no other deposit. Of the 26 claims, 19 cannot be tested.
Selection not documentedThe data set holds 200 mutant and 122 control units (and 3 from a Bax -/- mutant), where the paper's Figure 9 has 115 mutant and 143 control cells. The data set does not say how its units relate to the paper's, which were kept, or why the counts differ.
Paper inconsistentThe data set labels 95 of its 122 control units "WT (chx10 het)", while the paper's Figure 9 legend gives the controls as Pax6-Cre;Pcdh-gamma +/f-del or Pcdh-gamma f-del/f-del without Cre. The labels and the paper disagree on what the controls were.
File format errorIn the 8 recordings of 20080628_R4, 20080628_R6, 20080702_R1 and 20080702_R4, 82 of 86 spike trains are not in time order. Each runs to 1440 s and then holds a second segment between 720 and 780 s, though the stimulus runs to 1500 s. The 4 trains in order have spikes between 1440 and 1500 s, about as many as between 720 and 780 s, and moving the segments by 720 s raises their STA peaks over the null in most units. This is consistent with the last minute stored with a 720 s offset in place of 1440 s. One checkerboard recording (20080628_R5) has 1 or 2 spikes near 2 s out of order in 15 units.
Metadata errorThe frame lengths stated for the vertical-bar recordings of 20080702_R1 and 20080702_R4 (16.6691 and 16.6653 ms) are not those of the stimulus. Rebuilt with the stated frame, the bars of 20080702_R4 give no STA above the null in any of its 12 units (median ratio 0.96), though all 12 respond to the checkerboard of the same retina. A scan of the frame peaks at 16.671 to 16.672 ms in both recordings; with 16.6718 ms, the frame of every other bar recording and half the checkerboard frame of the same two retinas, the median ratio is 2.57 and 10 of the 12 units have peaks above twice the null's.
Other16 pairs of units share 70 to 100% of their spikes within 0.5 ms in every recording of their retina (12 pairs in mutant retinas, 4 in a control retina), most likely single neurons sorted twice. The data set does not mark them.
Description missingThe files do not give the age of the mice, the Pcdh-gamma allele (the paper's Results say fcon3 and its Figure 9 legend f-del), the region recorded in the Pax6-Cre retinas (the paper used their periphery only), or the cell type of the units.

Notebooks

Every registered entry that uses this dataset. An entry can hold several notebooks, and can use more than one dataset.

Reproduction Retinal ganglion cell responses in Pcdh-gamma mutant retina (Lefebvre et al. 2008)

Tests the findings of Lefebvre, Zhang, Meister, Wang and Sanes (2008) on CRCNS data set ret-1, which holds multielectrode recordings of 325 units from 16 mutant and control mouse retinas stimulated with flickering bars and checkerboards. It compares receptive field size, response time course, firing rate and Linear-Nonlinear model threshold and gain between Pcdh-gamma mutant and control retinas (Figure 9D to 9H). The histology of Figures 1 to 8 and the flash, spot and grating experiments of Figure 9A to 9C are not in the data set. All files

Reproducing Lefebvre et al. (2008), Figure 9, from CRCNS ret-1
reproduce_lefebvre_2008.ipynbExecuted, with outputs · Source

Verification

Round 1
2026-10-20Agreed with 26 of 26 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 1. The first executed notebook with the conclusion cells and the summary removed, the helper module and the claims file, with the paper. The verifier checked the archive's bucket listing and filelist.txt and tested helper functions on synthetic inputs. Its report gave its own outcome for each claim; all 26 agree with the record as it then stood. B1 to B9 in the reasons are the numbered problems of section B of its report (transcripts/crcns-ret-1/verifier-1-report.md); it marked B2 and B3 as affecting a conclusion and the others as not.
Round 2
2026-10-20Agreed with 26 of 26 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 2. The full executed notebook after the round-1 revision, with its conclusions and summary, the README, the helper module, the claims file, this record and the round-1 report, with the paper. It agrees with all 26 outcomes. Cell numbers in the reasons are those of the notebook it reviewed; in the revised notebook cells 0 to 54 keep their numbers, cells 55 to 59 are new (the variants under every reading) and the summary is cell 60.

Method differences

How the tests differed from the paper’s method or data, and the claims each applies to.

The vertical bars of 20080702_R1 and 20080702_R4 are rebuilt with a frame of 16.6718 ms, as for fig9d-rf-size-similar.s3-receptive-field-surrounds fig9e-slower-time-course fig9g-threshold-unaltered fig9h-gain-reduced
The claims file asks also for the count of units with a flicker response without defining one; the counts above two thresholds of the STA peak over a null STA are shown as a standard choice, and none can change the outcome.fig9-cell-counts
The control retinas are as the data set labels them; 95 of the 122 control units carry a label ("WT (chx10 het)") that the paper's description of its controls does not include.fig9-cell-counts
All 5 retinas recorded with 8.3 um bars are mutant; every control retina was recorded with 16.6 um bars.fig9d-rf-size-similar
The mutant group cannot be restricted to the peripheral region of the Pax6-Cre retinas, as the data set does not say where they were recorded.fig9d-rf-size-similar
The vertical bars of 20080702_R1 and 20080702_R4 are rebuilt with a frame of 16.6718 ms, the frame of the other bar recordings, not the stated 16.6691 and 16.6653 ms (see findings).fig9d-rf-size-similar
Lag j is placed at (j + 0.5) frames, the data set's example script's convention; the offset is the same for every unit.fig9e-slower-time-course
Theta is searched on 99 quantiles of y for the frame fit and at every split between the 20 bin centres for the binned fit, a standard choice of resolution; the binned grid cannot place theta above the last split, which many printed-form fits reach.fig9g-threshold-unaltered

Provenance

Attempt record: crcns-ret-1-lefebvre-2008-01
Paper: 10.1242/dev.027912
Date: 2026-10-20
Made by: An agent, run by CatalystNeuro
Dataset: CRCNS:ret-1
Version: unversioned (doi:10.6080/K0RF5RZT)
Run on: not recorded
Harness: Claude Code (GitHub Actions)
Model: claude-opus-5-5
Protocol: pilot-10

Transcripts of the agent runs

Each run’s prompt, messages and tool calls, with long output cut and the text of the paper left out, and its final report.

DANDI:000003 →