Next generation CANCAN focusing for remote stimulation by nanosecond electric pulses (DANDI:000686)

Claims

From the paper Next generation CANCAN focusing for remote stimulation by nanosecond electric pulses

Reproducing Pakhomov et al. (2023) from dandiset 000686 4

Notebook reproduce_pakhomov_2023.ipynb

Not reproduced4

Not reproducedResults 3.2, paragraph 1; Figure 5Averifier disagreed at first
Without rotation, the control of 40 unipolar trains from e1 causes much stronger electroporation throughout the array than 40 Packet 2 packets started from e1. The uptake after trains follows the field strength distribution and falls with distance from e1.
The lack of cancellation caused a much stronger electroporation throughout the array, and the distribution of YP uptake expectedly followed the electric field strength distribution in Fig. 3C .
Results 3.2, paragraph 1; Figure 5A
Under all 8 readings (O1 centre ROI or LOESS at spans 0.1, 0.2, 0.3; O4 raw or minus the footprint background), n = 4 and 4 dishes, and under each of the six ROI geometries, the criterion fails, because at the e3 end the trains do not exceed the packets (main geometry 1138 against 1299 a.u., p = 0.558; 839 against 1068 after background, p = 0.373; other geometries p = 0.268 to 0.858). In the main geometry (ROIs one step wide, ending at the measured footprint) the trains exceed the packets in the centre under the 6 LOESS readings (p = 0.00332 to 0.0156), not with the centre ROI (p = 0.0657 and 0.069), and near e1 only in the mean (13282 against 6339 a.u., p = 0.0831); with ROIs 3 or 5 steps wide or the needle reading of the ROI end the e1-end difference holds in all 8 readings (p = 0.000246 to 0.0247 over those five geometries). The train e1-end value exceeds the train centre under all 8 main readings (unpaired p = 0.0344 to 0.0428) and in every geometry, and the train mean profile falls from 13282 to 17953 a.u. over the 2 mm nearest e1 to 789 to 1138 a.u. near e3. So the statement's second sentence holds, and its first holds in the centre and near e1, not at the far end of the diagonal. In the train dishes the e3 footprint lies in the dark half of the image, where a faint dark disc is visible under the e3 circle in all four dishes (image figure of the data check); the e3-end values are 764 to 1391 a.u. per dish at widths 1, 3 and 5.
How the test differed from the paper’s method
- The paper's ROI measurements are not deposited. Profiles are measured here from the images with automatically found electrode footprints and the ROI layout of Methods 2.4 (56 ROIs per half diagonal, from 224 along both diagonals, plus one on the centre, ending at 3.65/4.2 of the distance from the centre to the electrode's inner edge, from the +/- 3.65 mm of Methods 2.4 and the 8.4-mm diagonal of Results 3.5). Each ROI is one step (about 65 um) long along the diagonal, so that the ROIs tile it. The paper gives neither the ROI width across the diagonal nor how the inner edge was placed, so six geometries are run, widths of 1, 3 and 5 steps, and the inner edge at the footprint edge measured in the image or at 4.2/(4.2 + 0.75) of the centre-to-centre distance (1.5-mm needles, Methods 2.3). The main tables use width 1 and the measured footprint. The outcome is the same under all six.
- e1 is not labelled in the files. It is taken as the top-right footprint of the image, which is both the electrode with the brightest outermost ROIs in all 8 non-rotated dishes and the mirror image of the top-left electrode energized in Fig. 3 (an inverted microscope images the dish from below). The two readings agree in every dish.
- Background for O4 (b) is the median inside the electrode footprints (unexposed monolayer under the needle), an assumption; the paper names no background subtraction. No outcome changes with it.
Not reproducedResults 3.2, paragraph 1; Figure 5A
Without rotation, Packet 2 enhances electroporation in the center of the array despite the weakening of the field with distance from e1. Uptake near e1 still stays stronger than in the center.
In contrast, Packet 2 enhanced electroporation in the center of the array, despite gradual weakening of the electric field with distance from e1. However, the bipolar cancellation only partially offset the electric field weakening, and electroporation near e1 remained stronger than in the center.
Results 3.2, paragraph 1; Figure 5A
The e1-end value of the 40-packet e1 to e3 profile exceeds the centre in the mean (6339 against 2576 a.u., raw ROIs, main geometry) and in all 4 dishes (ratios 1.17 to 4.22), but under none of the 16 readings (O1 x O4 x O6) is the difference significant (p = 0.137 to 0.195 in the main geometry, 0.0533 to 0.231 over all six ROI geometries), and by the paper's error-bar rule it is not either (gap 0.37 to 0.66 s.e.). The centre exceeds the trough between the centre and e1 (found 1.60 to 1.80 mm from the centre) under 10 of 16 main readings, every LOESS reading unpaired and spans 0.1 and 0.2 paired (p = 0.00428 to 0.0357), but not with the centre ROI (p = 0.255 to 0.325) or span 0.3 paired (p = 0.0598); over the six geometries under 10 to 14 of 16, and under 10 to 14 with the trough found leave one out. The criterion needs both comparisons; the second fails under every reading and geometry, so all of them give the same outcome. The direction of both effects is as the paper states; with n = 4 dishes and this spread between dishes the e1 to centre difference is not significant.
How the test differed from the paper’s method
- Profiles measured from the images as in fig5a-trains-stronger (ROI layout and the six geometries, e1, background).
- The trough is the lowest point of the group's mean profile (ROI values, or the mean of the dishes' LOESS fits under O1 b) strictly between the centre and the e1 end; each dish's value at that ROI enters the test. The claims file says "the lowest value between the center and the e1 end" without saying whether per dish or on the mean; the mean profile avoids picking each dish's noise minimum. Choosing it on the dishes it is tested on favours centre > trough, so a variant finds each dish's trough on the mean of the other dishes (leave one out).
Not reproducedResults 3.2, paragraph 2; Figure 5Bverifier disagreed at firstquestion to the reader
With rotation (1P x 40), Packet 2 smears the effect at the perimeter and brings out a remote effect in the center. The same rotation of unipolar trains (1T x 40) leaves a void of electroporation in the center.
Rotation with Packet 2 “smeared” the electroporation effect at the perimeter and emphasized the remote effect in the center. This remote effect contrasted with a void of electroporation in the center when the pulse train was applied in the same manner instead of the packet.
Results 3.2, paragraph 2; Figure 5B
Profiles averaged over both diagonals by signed distance, n = 5 and 5. No ROI geometry gives both comparisons under every reading of O1, O2, O4 and O6. In the main geometry the 1P x 40 centre exceeds its lowest value (2.44 to 2.71 mm from the centre) in 24 of 32 readings, every LOESS reading (p = 0.00494 to 0.0327), not with the centre ROI (3581 against 1032 a.u., p = 0.0647 unpaired, 0.114 paired); the 1T x 40 centre is below its most peripheral value (centre 4884 to 5781 a.u. raw and 4570 to 5467 after background, against 10910 to 10989 and 10596 to 10675) in 20 of 32, every unpaired test (p = 0.0145 to 0.0347) and the paired centre-ROI tests (p = 0.0454 and 0.0482), not the paired LOESS tests (p = 0.0582 to 0.0879). Both hold in 12 of 32. With ROIs 3 or 5 steps wide both comparisons hold under all 16 unpaired readings (largest p 0.0329 for 1T x 40 and 0.0148 for 1P x 40), but the paired 1T x 40 test holds in 0 to 10 of 16 (largest p 0.0655 to 0.0904). Both differences go the way the paper says under every reading and geometry. The 1P x 40 centre-ROI mean rests mostly on dish2 (centre 7941 against ends 1575 a.u.; the other four dishes 1827 to 4280); in a full-resolution crop its centre ROI has a mean of 8151 a.u. with no saturated pixel, and pixels above 3 times the crop median cover 44.6% of it in two connected regions, the larger 49 um across (1.4 to 8.2% in the other dishes); by eye these are clusters of round cell-sized spots, consistent with labelled cells, though the measure does not exclude debris. The 1T x 40 centre is the lowest point of its ROI mean profile (4884 a.u.) but not of its LOESS fits; the criterion reports this and does not require it. This outcome reads the criterion as requiring both the paired and the unpaired test of O6 (the criterion cites O6, which runs both for comparisons inside the same dishes, and the fig5a-packet-center-enhanced criterion names both outright). Under the reading in which only the unpaired test counts, the outcome is the same with ROIs one step wide (12 of 16 unpaired readings; the 1P x 40 centre ROI fails, p = 0.0643 to 0.0693 under either ROI end), and both comparisons hold with ROIs 3 or 5 steps wide. ROIs one step wide are the most direct reading of the layout of Methods 2.4 (see fig5c-normalized-contrast), so under that width the outcome does not depend on the reading of O6. A question on the reading of O6 is in questions/000686.yaml.
How the test differed from the paper’s method
- Reading of the criterion. Both the paired and the unpaired test of O6 must hold; see the notes for the outcome under the other reading.
- The positive end of the averaged profile joins the e1 end and the e2 end, e2 taken as the footprint next to e1 clockwise in the image; with rotation every electrode starts the same number of packets, so O2 (b) is read here as the outermost ROI at that joined end (e1 and e2), a labelling choice. It gives the same outcome as O2 (a).
- Trough as in fig5a-packet-center-enhanced (group mean profile, with a leave-one-out variant).
- Profiles measured from the images as in fig5a-trains-stronger (six ROI geometries).
Not reproducedResults 3.2, paragraph 2; Figure 5Cverifier disagreed at firstrests on an assumptionquestion to the reader
Normalized to the most peripheral ROI (100%), rotated Packet 2 and rotated trains show sharply different spatial distributions. The packet's center is relatively high and the trains' center is relatively low.
The striking difference in the spatial distribution of electroporation between the CANCAN and control treatments is appreciated best by comparing the normalized fluorescence data, when the emission intensity in the most peripheral (closest to the electrode) ROI is taken as 100% ( Fig. 5C ).
Results 3.2, paragraph 2; Figure 5C
The 1P x 40 normalized centre is above 100% under all 32 readings in every ROI geometry (128 to 267% in the main one) and the 1T x 40 one below (43 to 64%), so the direction holds everywhere. The criterion requires 1P x 40 to exceed 1T x 40 at p < 0.05 under every reading. With ROIs one step wide, the assumed width, the difference is significant in 27 of 32 readings under either reading of the ROI end; the 5 that fail use the centre ROI (p = 0.0529 to 0.0601 with the measured footprint, 0.0512 to 0.0824 with the needle reading), so the claim is not reproduced, a near miss. The width decides the outcome. With ROIs three steps wide the test holds in all 32 readings under both ROI ends (largest p 0.0376 and 0.0489); with five steps in 31 of 32 (one at p = 0.0505, measured footprint) or all 32 (largest p 0.0331, needle reading). The question of the width is put to the paper's authors (findings). The claim depends on fig5b-rotation-center-peak, which is not reproduced.
How the test differed from the paper’s method
- Assumption. ROIs one step wide. Methods 2.4 says only that the measurements used "92 or 224 regions of interest (ROI), placed along one or both diagonals", which fixes the ROI length along the diagonal (2 x 3.65 mm / 112, about 65 um) but not its width. Square ROIs that tile the diagonal at that spacing without gaps or overlap are the most direct reading of that layout, and the main geometry of the notebook since its first version. ROIs 3 and 5 steps wide are shown as variants.
- O2 (a) and (b) as in fig5b-rotation-center-peak (O2 b is the joined e1 and e2 end).
- Assumption. For the test under O3 (b) each dish's centre is divided by its group's mean periphery; the claims file gives the reported value (mean profile over its own peripheral value) but not the per-dish values a test needs.
- Profiles measured from the images as in fig5a-trains-stronger (six ROI geometries).

Not tested in a notebook 16

16 data missing from the dataset

Blocked16

BlockedMethods 2.3, paragraph 4; Figure 3· Data missing from the dataset
In the simulations, 1 kV applied to one electrode gives about 1.8 kV/cm in the center of the small array and 0.8 kV/cm in the center of the large array. The field near the energized electrode is about 4-fold higher (7 and 3.5 kV/cm per kV). Energizing one diagonal gives no field in the center.
For example, applying 1 kV to one electrode produced the electric field of approximately 1.8 kV/cm and 0.8 kV/cm in center of the small and large arrays ... The electric field strength was about 4-fold higher near the energized electrode (7 kV/cm and 3.5 kV/cm per 1 kV applied, for the small and large arrays, respectively) Note that energizing either electrode diagonal produced no electric field in the center of the array
Methods 2.3, paragraph 4; Figure 3
The values are outputs of the paper's Sim4Life model. Dandiset 000686 holds only the 18 monolayer images of Fig. 5, and none of the four other dandisets that name the paper (000875, 001025, 001027, 001028) holds field maps; the paper names no deposit of the model. The criterion rules out a model built by the author.
BlockedResults 3.1, paragraph 2; Figure 4· Data missing from the dataset
Sham-exposed cells gain no YP fluorescence over the 10 min of imaging. Cells given pulses show a gradual rise in YP signal from the first frame after exposure.
Sham-exposed cells showed no gain in YP fluorescence during the 10 min of observation. In contrast, cells treated by electric pulses displayed gradual increase in YP signal, starting from the image frame taken immediately after nsEP exposure.
Results 3.1, paragraph 2; Figure 4
Needs the confocal time-lapse of single cells on coverslips (Methods 2.2). Not in dandiset 000686 (one stitched monolayer image per file), not in the four other dandisets that name the paper (their descriptions are monolayer experiments of Figs. 6 to 9), and not found in a search of DANDI for the lab's other dandisets (search terms CANCAN, Pakhomov, YO-PRO, electroporation); the paper names no deposit.
BlockedResults 3.1, paragraph 2; Figure 4· Data missing from the dataset
A single 600-ns pulse at 15 kV/cm gives the strongest YP uptake, 126 +/- 14 a.u. (n = 25 cells) by the end of recording.
A single 600-ns pulse at 15 kV/cm caused by far the strongest YP uptake, reaching 126±14 a.u. (n=25) by the end of recording.
Results 3.1, paragraph 2; Figure 4
Single-cell time-lapse data not deposited; see fig4-sham-no-uptake.
BlockedResults 3.1, paragraph 2; Figure 4· Data missing from the dataset
Bipolar signals that start with the same pulse cause much weaker uptake (p < 0.001). By the end of recording the 4-phasic signal reaches 20.8 +/- 2.0 a.u., the 8-phasic 6.9 +/- 1.6 a.u., and the 8-phasic with 5-us delays 25.4 +/- 2.8 a.u.
Bipolar signals started with exactly the same pulse (600 ns, 15 kV/cm) but caused significantly weaker effects (p<0.001). By the end of recording, YP signal reached only 20.8±2.0 a.u. for the 4-phasic signal; 6.9±1.6 a.u. for the 8-phasic signal; and 25.4±2.8 a.u. for the 8-phasic signal with 5-μs delays between phases.
Results 3.1, paragraph 2; Figure 4
What the paper reports
- each bipolar signal against the single unipolar pulse: two-sided Student's t-test (the test named in Methods 2.4), p < 0.001 (n = 21 to 30 cells per group)
Single-cell time-lapse data not deposited; see fig4-sham-no-uptake.
BlockedFigure 4 legend; Results 3.1, paragraph 2; Results 3.4, paragraph 1· Data missing from the dataset
The 8-phasic signal with no delays causes about 20-fold weaker electroporation than a single unipolar pulse. The 4-phasic signal and the delayed 8-phasic signal cancel 4 to 5-fold.
Note that applying the 8-phasic bipolar signal with no inter-phase delays caused about 20-fold weaker electroporation than a single unipolar pulse
Figure 4 legend; Results 3.1, paragraph 2; Results 3.4, paragraph 1
Depends on fig4-unipolar-uptake and fig4-bipolar-weaker, both blocked for missing data.
BlockedResults 3.3, paragraph 1; Figure 6A· Data missing from the datasetquestion to the reader
Raising the total number of rotated Packet 2 packets from 4 to 20 and 40 strongly raises YP uptake throughout the array. A central peak is discernible for every treatment.
Increasing the total number of applied packets from 4 to 20 and 40 profoundly enhanced the electroporative YP dye uptake throughout the electrode array ( Fig. 6A ). The peak of electroporation in the center was clearly discernable for all tested treatments.
Results 3.3, paragraph 1; Figure 6A
The Fig. 6 images (1Px4, 5Px4 and 10Px4) are not in dandiset 000686, the dataset the claims file names for this claim (data dataset), so outside data may not be used (protocol, author, step 4). They appear to be in DANDI:000875 (version 0.240515.1925, 9 files, "Influence of unipolar pulse packet quantity on centering the electroporation effect within a quadrupole electrode array"), which names this paper as IsPublishedIn; that dandiset was not opened. A question in questions/000686.yaml asks a reader whether the claim should be data elsewhere.
BlockedResults 3.3, paragraph 1; Figure 6B· Data missing from the dataset
The center is almost twofold brighter than the periphery for 1Px4. It is about the same as the periphery for 5Px4 and 10Px4.
Its brightness was almost twofold higher than at the periphery for a single packet was applied 4 times from different directions, or about the same as at the periphery for the rotation of 5 and 10 packets ( Fig. 6B ).
Results 3.3, paragraph 1; Figure 6B
Fig. 6 data are not in 000686 (apparently in DANDI:000875); see fig6a-more-packets-more-uptake.
BlockedResults 3.3, paragraph 1; Figure 6· Data missing from the dataset
For all three packet numbers, the least affected area (between the center and the electrodes) shows about 2-fold lower fluorescence than the center.
Interestingly, the least affected area of the cell monolayer (between the center and the electrodes) showed about 2-fold lower fluorescence intensity than in the center for all the treatments.
Results 3.3, paragraph 1; Figure 6
Fig. 6 data are not in 000686 (apparently in DANDI:000875); see fig6a-more-packets-more-uptake.
BlockedResults 3.3, paragraph 1; Figure 6· Data missing from the dataset
The central peak becomes narrower with more packets.
Also, the central peak appeared narrower (sharper) with increasing the number of packets.
Results 3.3, paragraph 1; Figure 6
Fig. 6 data are not in 000686 (apparently in DANDI:000875); see fig6a-more-packets-more-uptake.
BlockedResults 3.3, paragraph 2; Figure 7A· Data missing from the dataset
Raising the first-pulse amplitude from 2.4 to 5.6 kV (1Px4, small array) raises YP fluorescence about equally everywhere. It rises 15.5 times at the most peripheral point, 16.2 times half-way to the center and 16.8 times in the center.
For example, when the first pulse amplitude was increased from 2.4 to 5.6 kV ... YP fluorescence increased 15.5 times at the most peripheral point; 16.2 times half-way to the center; and 16.8 times in the center ( Fig. 7A ).
Results 3.3, paragraph 2; Figure 7A
The Fig. 7 images are not in dandiset 000686, which holds only 6.4-kV dishes. They appear to be in DANDI:001025 (version 0.240605.1826, 21 files, voltage and array size), which names this paper; see fig6a-more-packets-more-uptake for why they are not used.
BlockedResults 3.3, paragraph 2; Figures 7B, 7C and 7D; Discussion, paragraph 5· Data missing from the dataset
Normalized to the most peripheral ROI, profiles for different amplitudes fully overlap. The amplitude changes the degree of electroporation but not its spatial distribution. This holds for rotation of multiple packets (10Px4) and for the larger array.
When the YP emission was normalized to the most peripheral ROI, the data for different nsEP amplitudes fully overlapped ( Fig. 7B ). ... The findings held true for rotating multiple packets and when using the larger array ( Fig. 7C and D ).
Results 3.3, paragraph 2; Figures 7B, 7C and 7D; Discussion, paragraph 5
Fig. 7 data are not in 000686 (apparently in DANDI:001025); see fig7a-amplitude-scales-uniformly.
BlockedResults 3.4, paragraph 2; Figure 8A· Data missing from the dataset
Packets 1, 2 and 3 at 4 kV, rotated one cycle, give the same YP uptake in the center (215 to 235 a.u.).
All packets produced the same YP uptake in the center (215–235 a.u.)
Results 3.4, paragraph 2; Figure 8A
The Fig. 8 images are not in dandiset 000686, which holds only Packet 2 and its train. They appear to be in DANDI:001027 (version 0.240605.1859, 10 files, three protocols at 4 kV), which names this paper; see fig6a-more-packets-more-uptake for why they are not used.
BlockedResults 3.4, paragraph 2; Figure 8A· Data missing from the dataset
Packets 1 and 3 cause much more YP uptake near the electrodes (250 to 350 a.u.) than Packet 2 (about 120 a.u.), p < 0.01.
but Packets 1 and 3 caused drastically more YP uptake near the electrodes (250–350 a.u. vs ~120 a.u. for packet 1, p<0.01, Fig. 8A )
Results 3.4, paragraph 2; Figure 8A
What the paper reports
- Packets 1 and 3 against Packet 2 near the electrodes: two-sided Student's t-test (the test named in Methods 2.4), p < 0.01 (n = 3 to 4 experiments per packet)
Fig. 8 data are not in 000686 (apparently in DANDI:001027); see fig8-center-equal-packets.
BlockedResults 3.4, paragraph 2; Figure 8B· Data missing from the dataset
The normalized center YP emission is 183% of the periphery for Packet 2, but 83% for Packet 3 and 74% for Packet 1.
The superiority of Packet 2 for CANCAN targeting is highlighted by the drop of normalized YP emission in the center from 183% for Packet 2 to 83% and 74% for Packets 3 and 1, respectively ( Fig. 8B ).
Results 3.4, paragraph 2; Figure 8B
Fig. 8 data are not in 000686 (apparently in DANDI:001027); see fig8-center-equal-packets.
BlockedResults 3.5, paragraph 1; Figure 9A· Data missing from the dataset
With the large array lifted 1, 2 and 3 mm above the monolayer, YP emission near the electrodes falls 2.4, 6.4 and 9 times, against electrodes touching the cells. In the center it falls only 1.02, 1.4 and 3.8 times.
Specifically, YP emission near the electrodes decreased 2.4, 6.4, and 9 times with 1-, 2- and 3-mm gaps (as compared to electrodes being in the same plane with cells); in the center, it decreased just 1.02, 1.4, and 3.8 times, respectively.
Results 3.5, paragraph 1; Figure 9A
The Fig. 9 images are not in dandiset 000686, which has no electrode-gap conditions. They appear to be in DANDI:001028 (version 0.240605.1954, 12 files, electrodes 0 to 3 mm above the monolayer), which names this paper; see fig6a-more-packets-more-uptake for why they are not used.
BlockedResults 3.5, paragraph 1; Figures 9B and 9C· Data missing from the dataset
The faster fading of the effect near the electrodes raises targeting efficiency (center over periphery) to as much as 600% for the 2-mm gap. In the images, electroporation is confined almost entirely to the center when the electrodes are above the monolayer.
Such disproportionally faster fading of the effect near electrodes boosted the CANCAN targeting efficiency (measured as a ratio of the effect in the center and at the periphery) to as much as 600% for the 2-mm gap ( Fig. 9B ).
Results 3.5, paragraph 1; Figures 9B and 9C
Fig. 9 data are not in 000686 (apparently in DANDI:001028); see fig9a-gap-reduces-periphery-more.

Findings about the data

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

Data missingThe dandiset holds only the 18 monolayer images of Fig. 5 (40 Packets and 40 Trains, 4 dishes each; 1P x 40 and 1T x 40, 5 dishes each). The single-cell time-lapse data of Fig. 4 and the field simulations of Fig. 3 are not in it and were not found in any other dandiset.
OtherThe paper's monolayer data are split over five dandisets (000686 for Fig. 5, 000875, 001025, 001027 and 001028 for Figs. 6 to 9), each naming the paper, but none names the others and the paper names none of them, so a reader of one cannot find the rest.
OtherThe files hold the stitched images only, not the ROI measurements (positions and intensities) that the paper's profiles plot, so the figures can be rebuilt only by placing ROIs anew.
Metadata errorThe species is NCBITaxon_9913 (Bos taurus) in every file and in the dandiset summary, where the paper used CHO-K1 cells (Chinese hamster, Cricetulus griseus); the strain field does say "cellline CHO". The subject age is P1D, which has no meaning for a cell line.
Metadata errorThe dandiset's assets summary gives 36 subjects, but the 18 files have 18 distinct subject ids, one per dish.
Metadata errorThe image description says the 144 tiles were captured "from a six-well plate", where the paper (Methods 2.1) seeded the monolayers in 35-mm glass-bottom dishes, one dish per experiment.
Description missingThe image has no pixel size (resolution -1, unit n.a.). The tile geometry (22323 pixels from 12 tiles with about 10% overlap of 2048-pixel ORCA-Flash4 frames under a 10x objective) is consistent with 0.65 um per pixel; at that scale the footprint centres are 9.7 to 10.1 mm apart across the diagonal and their inner edges 8.4 to 8.8 mm, consistent with the paper's 8.4-mm diagonal if that is measured between inner edges. The footprints are 1.15 to 1.44 mm across, against the 1.5-mm needles of Methods 2.3.
Description missingThe electrode footprints are not labelled, so which one is e1 (needed for Fig. 5A) is not stated. In all 8 non-rotated dishes the top-right footprint of the image has the brightest outermost ROIs, consistent with e1 there.
Methods not availableThe paper does not say how the centre value of a profile is taken, the ROI at distance 0 or the LOESS fit there. With ROIs one step wide, the 1P x 40 centre exceeds its trough (Fig. 5B) with the LOESS value (p = 0.0049 to 0.033) but not with the single centre ROI (p = 0.064 to 0.114). Which was used, and with what span?
Methods not availableFor comparisons inside the same dishes (centre against periphery, Fig. 5B), the paper does not say whether its two-sided Student's t-test is paired or unpaired. For 1T x 40 the unpaired test gives p = 0.015 to 0.035 under every reading and the paired one up to p = 0.09. Which test was used?
Methods not availableFor normalized profiles averaged over both diagonals (Fig. 5C), the paper does not say whether "the most peripheral ROI" is the mean of the outermost ROIs at both ends or one end, nor whether each dish is normalized before averaging. Which ROI is taken as 100%, and is each dish normalized first?
Methods not availableThe paper gives neither the size of its ROIs across the diagonal, nor how close to the electrode the outermost ROI sits, nor any background subtraction for the monolayer images. Background changes no outcome here. ROI width decides fig5c-normalized-contrast, whose test holds under every reading with ROIs 3 steps (about 195 um) wide and fails in 5 of 32 readings (p = 0.051 to 0.082) with ROIs 1 step (about 65 um) wide; it also moves the Fig. 5A trains against packets comparison near e1 from p = 0.083 (1 step) to 0.015 to 0.024 (3 and 5 steps). What ROI size, placement and background were used?

Questions

Where the author or a verifier believed the claims file misread the paper, they asked, citing the paper, and a reader who had not seen the data answered from the paper alone, amending the claims file where it was wrong. The author then ran the claim as answered. Attempts before protocol pilot-11 also asked here for method details the claims file did not give.

openasked 2026-10-31about fig6a-more-packets-more-uptake
The monolayer data of Figs. 6 to 9 are not in dandiset 000686, which holds only the 18 images of Fig. 5. Four other published dandisets of the same lab name this paper as the one they are published in (relatedResource IsPublishedIn 10.1016/j.bioelechem.2023.108437) and their titles and descriptions match the figures, DANDI:000875 (0.240515.1925, packet number, Fig. 6), DANDI:001025 (0.240605.1826, voltage and array size, Fig. 7), DANDI:001027 (0.240605.1859, the three packets at 4 kV, Fig. 8) and DANDI:001028 (0.240605.1954, electrodes 0 to 3 mm above the monolayer, Fig. 9). Should the claims on Figs. 6 to 9 (fig6a-more-packets-more-uptake, fig6b-center-to-periphery, fig6-void-between, fig6-peak-narrower, fig7a-amplitude-scales-uniformly, fig7-amplitude-no-effect-on-distribution, fig8-center-equal-packets, fig8a-periphery-packets-1-3-higher, fig8b-normalized-center, fig9a-gap-reduces-periphery-more, fig9b-gap-boosts-targeting) say `data` set to `elsewhere` with these dandisets under `where`, so that an attempt may test them there?
Passage. Each dish was used in one experiment (Methods 2.1). Fluorescence data are presented as a mean +/- s.e. for 3-5 independent experiments (Methods 2.4). The claims file's header says "Which of the other experiments it holds is for the author to find", and every claim has the default `data` set to `dataset`.
Why it matters. With `data` set to `dataset` these eleven claims cannot use outside data (protocol, author, step 4) and are blocked here, although the paper's own deposits of their data are public and pinned. Marked `elsewhere`, they could be tested on those dandisets in an `author-answers` run.
No reader has answered it yet.
openasked 2026-10-31about fig5b-rotation-center-peak
The criterion requires both comparisons "under all readings of O1, O2 and O4" and says "Each comparison uses a two-sided t-test at p < 0.05 (Methods 2.4, O6)". O6 runs both the paired and the unpaired test for comparisons inside the same dishes. Must both tests hold for the claim to be reproduced, as the fig5a-packet-center-enhanced criterion says outright ("paired and unpaired, O6"), or does one of them suffice, and if so which?
Passage. Methods 2.4 says only "Two-sided Student's t-test was employed to analyze the significance of differences; p<0.05 was considered statistically significant", without paired or unpaired. The claims file's O6 says "For comparisons inside the same dishes (center against periphery) both are run."
Why it matters. With ROIs 3 or 5 steps wide every unpaired reading holds and only the paired 1T x 40 test fails (p up to 0.0655 to 0.0904), so the outcome there turns on this reading. The attempt reads the criterion as requiring both; with ROIs one step wide it is not reproduced under either reading, because the 1P x 40 centre ROI fails the unpaired test too (p = 0.0643 to 0.0693). The round-2 verifier reported the criterion as ambiguous on this point.
No reader has answered it yet.
openasked 2026-10-31about fig5c-normalized-contrast
O1 says "The paper gives no ROI size". The paper's ROI count and range fix the ROI length along the diagonal (2 x 3.65 mm / 112, about 65 um) and leave only the width across the diagonal open. Should O1 say so, and should it list the ROI width as a reading, since the fig5c-normalized-contrast outcome turns on it?
Passage. Methods 2.4 says the measurements used "92 or 224 regions of interest (ROI), placed along one or both diagonals"; the claims file's header places them over +/- 3.65 mm from the array centre for the large array, and its O1 says "The paper gives no ROI size or window".
Why it matters. The round-3 verifier reported the wording of O1 as a misreading. With ROIs one step (about 65 um) wide the fig5c test fails in 5 of 32 readings and the claim is not reproduced; with ROIs 3 steps wide it holds in all 32. The width is not one of the claims file's readings, so the attempt took one step as an assumption.
No reader has answered it yet.

Notebooks

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

Reproduction Next generation CANCAN focusing of electroporation (Pakhomov et al. 2023)

Tests the Fig. 5 findings of Pakhomov, Gudvangen, Mangalanathan, Kondratiev, Redondo and Semenov (2023) on dandiset 000686, which holds 18 stitched YO-PRO-1 images of CHO cell monolayers in a quadrupole electrode array after 40 Packet 2 packets or 40 control trains, with and without rotation. The profiles along the array diagonals are measured from the images. The single-cell data of Fig. 4, the simulations of Fig. 3 and the monolayer data of Figs. 6 to 9 are not in this dandiset. All files

Reproducing Pakhomov et al. (2023) from dandiset 000686
reproduce_pakhomov_2023.ipynbExecuted, with outputs · Source

Verification

Round 1
2026-10-31
Agreed with 17 of 20 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 queried the archive's asset list and read the helpers without running them (no numpy or scipy in its environment). It agreed with the 5 Fig. 3 and 4 blocks and with fig5a-packet-center-enhanced, gave not_reproduced for the other three Fig. 5 claims, and gave data_missing for the 11 Fig. 6 to 9 blocks.
- fig5a-trains-stronger: the verifier’s outcome was “Not reproduced”. Fails under all 8 readings. At the e3 end trains are not above packets (1099 vs 1285 a.u., p=0.52). At the e1 end the difference misses p<0.05 (13565 vs 6347, p=0.059). Centre (p 0.004 to 0.044) and train e1 > centre (p about 0.03) hold. All readings agree, so it is not methods_underspecified. It asked for the per-dish values at width 3 (C).
- fig5b-rotation-center-peak: the verifier’s outcome was “Not reproduced”. Both parts hold in only 12 of 32 readings. Under the centre-ROI reading, 1P x 40 centre > trough has p=0.073 unpaired and 0.125 paired. Under the paired LOESS readings, 1T x 40 periphery > centre has p=0.05 to 0.08. All directions are as claimed, and the criterion requires every reading. ROI width is open in the paper, changes outcomes, and was run only for the main reading with overlapping squares; a band one step long and wider across the diagonal would be the consistent variant (B2). The ROI end depends on a footprint radius estimated from intensity (B4). Whether the 1P x 40 dish2 centre (8148) is cells or debris was not checked (C).
- fig5c-normalized-contrast: the verifier’s outcome was “Not reproduced”. Holds in 24 of 32 readings. 7 of 8 centre-ROI readings have p=0.051 to 0.057, and one LOESS 0.3 reading has p=0.0515. 1P x 40 is above 100% in every reading (134 to 268%), against 42 to 63% for 1T x 40. A near miss in a consistent direction. With 3-step ROIs the contrast has p=0.014, so ROI width changes the outcome (B2). The O3 (b) test is the author's construction and is not labelled as an assumption (B6).
Round 2
2026-10-31Agreed with 20 of 20 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 helper module, the claims file, the README and the attempt record, with the paper. The verifier agreed with all 20 outcomes, with caveats on fig5b-rotation-center-peak and fig5c-normalized-contrast, and quoted sentences the printed outputs did not support. In reply to points not tied to one claim, the data check now prints the subject age (P1D) and the dandiset's assets summary (numberOfSubjects 36, 18 files, 18 distinct subject ids), which the summary and the findings cite; the README says the Figs. 6 to 9 data appear to be in four other dandisets, not opened here; the finding on the centre ROI gives p = 0.064 (the lowest printed, 0.0643); and a new notebook section checks the helpers on a synthetic image (footprints found within 14 blocks of the true centres, radius 99 against 100, background 200 against 200, ROI means within 2 a.u. of a known ramp, LOESS, t-tests and the error-bar rule against values computed by hand).
Round 3
2026-10-31Agreed with 1 of 1 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 3. The one proposed change of outcome, on fig5c-normalized-contrast, and the analysis behind it in the full executed notebook (the section "Proposed change, the ROI width of Methods 2.4", cells 21 and 24), with the helper module, the claims file and the paper. Nothing else was reviewed.

Method differences

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

Profiles measured from the images as in fig5a-trains-stronger (six ROI geometries).fig5b-rotation-center-peak fig5c-normalized-contrast
The paper's ROI measurements are not deposited. Profiles are measured here from the images with automatically found electrode footprints and the ROI layout of Methods 2.4 (56 ROIs per half diagonal, from 224 along both diagonals, plus one on the centre, ending at 3.65/4.2 of the distance from the centre to the electrode's inner edge, from the +/- 3.65 mm of Methods 2.4 and the 8.4-mm diagonal of Results 3.5). Each ROI is one step (about 65 um) long along the diagonal, so that the ROIs tile it. The paper gives neither the ROI width across the diagonal nor how the inner edge was placed, so six geometries are run, widths of 1, 3 and 5 steps, and the inner edge at the footprint edge measured in the image or at 4.2/(4.2 + 0.75) of the centre-to-centre distance (1.5-mm needles, Methods 2.3). The main tables use width 1 and the measured footprint. The outcome is the same under all six.fig5a-trains-stronger
e1 is not labelled in the files. It is taken as the top-right footprint of the image, which is both the electrode with the brightest outermost ROIs in all 8 non-rotated dishes and the mirror image of the top-left electrode energized in Fig. 3 (an inverted microscope images the dish from below). The two readings agree in every dish.fig5a-trains-stronger
Background for O4 (b) is the median inside the electrode footprints (unexposed monolayer under the needle), an assumption; the paper names no background subtraction. No outcome changes with it.fig5a-trains-stronger
Profiles measured from the images as in fig5a-trains-stronger (ROI layout and the six geometries, e1, background).fig5a-packet-center-enhanced
The trough is the lowest point of the group's mean profile (ROI values, or the mean of the dishes' LOESS fits under O1 b) strictly between the centre and the e1 end; each dish's value at that ROI enters the test. The claims file says "the lowest value between the center and the e1 end" without saying whether per dish or on the mean; the mean profile avoids picking each dish's noise minimum. Choosing it on the dishes it is tested on favours centre > trough, so a variant finds each dish's trough on the mean of the other dishes (leave one out).fig5a-packet-center-enhanced
Reading of the criterion. Both the paired and the unpaired test of O6 must hold; see the notes for the outcome under the other reading.fig5b-rotation-center-peak
The positive end of the averaged profile joins the e1 end and the e2 end, e2 taken as the footprint next to e1 clockwise in the image; with rotation every electrode starts the same number of packets, so O2 (b) is read here as the outermost ROI at that joined end (e1 and e2), a labelling choice. It gives the same outcome as O2 (a).fig5b-rotation-center-peak
Trough as in fig5a-packet-center-enhanced (group mean profile, with a leave-one-out variant).fig5b-rotation-center-peak
Assumption. ROIs one step wide. Methods 2.4 says only that the measurements used "92 or 224 regions of interest (ROI), placed along one or both diagonals", which fixes the ROI length along the diagonal (2 x 3.65 mm / 112, about 65 um) but not its width. Square ROIs that tile the diagonal at that spacing without gaps or overlap are the most direct reading of that layout, and the main geometry of the notebook since its first version. ROIs 3 and 5 steps wide are shown as variants.fig5c-normalized-contrast
O2 (a) and (b) as in fig5b-rotation-center-peak (O2 b is the joined e1 and e2 end).fig5c-normalized-contrast
Assumption. For the test under O3 (b) each dish's centre is divided by its group's mean periphery; the claims file gives the reported value (mean profile over its own peripheral value) but not the per-dish values a test needs.fig5c-normalized-contrast

Provenance

Attempt record: 000686-pakhomov-2023-01
Paper: Next generation CANCAN focusing for remote stimulation by nanosecond electric pulses
Date: 2026-10-31
Made by: An agent, run by CatalystNeuro
Dataset: DANDI:000686
Version: 0.240515.1909 (doi:10.48324/dandi.000686/0.240515.1909)
Run on: GitHub Actions
Harness: Claude Code
Model: claude-opus-5-5
Protocol: pilot-12

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.

Other datasets that name the same paper

← DANDI:000678 DANDI:000688 →