23 rows: 9 dose_response (4 LC50 from Table 2, 3 AChE IC50 from Table 3, 2 Ki from §3.6) and 14 time_response (3 antifeedant rates from §3.1, 11 fraction mortalities from Table 1). Every compound potency and ladder is in µg/mL.
The abstract and conclusion restate the same values; they are not extra rows.
","page":1,"rects":[[36.78,39.33,92.21,75.21]]},{"id":"intro-stats","cat":"excl","crit":[],"title":"Introduction figures are not rows","body":"46 countries, 617 counties in 16 provinces, 126,700 ha, and the >170 known compounds are background. Emitting any of them as a row is over-extraction.
","page":2,"rects":[[28.29,23.33,36.49,24.76],[29.45,30.75,37.7,32.18],[29.81,32.24,40.61,33.67],[88.2,12.53,92.03,13.96]]},{"id":"isolation","cat":"crit","crit":[3,8],"title":"Isolation route and the Lx fractions","body":"EtOAc layer (372 g) → silica gel column 10 × 100 cm, PE–EtOAc 50:1 → pure EtOAc → Lx.1–Lx.12, the Table 1 fractions (mixtures, not pure compounds).
Lx.9 (18.9 g) → MCI → Lx.9G (800 mg) → silica → Lx.9G3 (200 mg) → gel column → prep HPLC (37% CH3CN, 0.1% FA) → 1: 117 mg, tR 15 min; 2: 71 mg, tR 37 min.
Layer masses (PE 253 g, EtOAc 372 g, n-BuOH 324 g) belong in descriptions, never as rows. Only Lx.9 has a stated mass; invented per-fraction eluent ratios or masses are penalized.
","page":3,"rects":[[7.89,56.55,49.38,88.4],[50.72,9.39,92.21,26.02]]},{"id":"plant","cat":"crit","crit":[3],"title":"Provenance for compounds and extracts","body":"Collected July 2022 at Anyang, Henan (35°12′N, 113°37′E), sample No. 2022017, identified by Dr Guoxing Wu. This belongs in the descriptions of compounds 1 and 2 and of the extracts and fractions, never as rows.
The paper says only that the collected plant was dried and pulverized, so “whole plant” is a minor embellishment.
","page":3,"rects":[[7.89,23.88,49.38,33.51]]},{"id":"antifeedant-method","cat":"crit","crit":[1],"title":"Antifeedant assay instructions","body":"Leaf-disc method (Tang and Wang 2007), extracts at 10 mg/mL (genuinely mg: do not convert), acetone control, discs d = 1.2 cm soaked 3 s, replaced once more than 80% is eaten, dishes at 25–30 °C, area eaten measured on grid paper, the rate formula, and the row's time point. n_samples = 3.
A 48 h reading is scheduled but never reported, so any 48 h row is penalized. The leaf species isn't stated; “corn leaf discs” is a minor embellishment.
","page":3,"rects":[[50.72,28.4,92.21,56.91]]},{"id":"rearing","cat":"crit","crit":[7],"title":"Larva description, used everywhere","body":"Every S. frugiperda ref (antifeedant, Table 1, both LC50 routes, and the AChE source) should say: reared for multiple generations on corn leaves at the Insect Toxicology Laboratory, Yunnan Agricultural University; 26 °C, 60% RH, 14 h:10 h; 4th-instar larvae. The duplicated paragraph is not a conflict, and no strain or field origin is given.
Facts only in shared_instructions with bare target descriptions: partial. A different instar, artificial-diet rearing, a field population, or the 25–30 °C / 37 °C assay temperatures given as rearing conditions: error.
","page":3,"rects":[[7.89,35.76,49.38,54.3]]},{"id":"dietary-method","cat":"crit","crit":[2,5],"title":"Dietary (“gastrotoxic”) route","body":"Artificial-feed film: 0.1 mL acetone solution spread on 1 g feed and dried in the shade, one 4th-instar larva per dish, 10 dishes, experiment run twice → n_samples = 2 on dietary LC50 rows (not 20). Solvent and blank controls.
The 200 µg/mL screen of Figure 3A is plot-only: no rows.
","page":3,"rects":[[50.72,59.76,92.21,88.4]]},{"id":"invivo-ache","cat":"excl","crit":[4],"title":"In-vivo AChE assay (Figure 5A)","body":"Fixed LC50 and 2×LC50 doses, plotted only: no rows. The “five replicates” belong to this assay, so n_samples = 5 on the in-vitro IC50 rows is wrong.
Immersion for 3–4 s, 10 dishes with one larva each, run three times → n_samples = 3 on contact LC50 rows (not 30). Solvent and blank controls. Swapping the two routes in Table 2 is an error.
Target: protein “acetylcholinesterase”, organism S. frugiperda, described as the crude supernatant of ~0.1 g 4th-instar larvae homogenized in 1 mL buffer (ice bath 5 min; 10,000 rpm, 4 °C, 10 min). A generic “larval homogenate” is a partial.
Instructions: 4 µL compound + 96 µL enzyme, final 3.75, 7.5, 15, 30, 60 µg/mL, 37 °C for 2 h; 50 µL 1.5 mM acetylcholine for 5 min; 50 µL 0.3 mM DTNB; read at 412 nm; inhibition = (ODc − ODt)/ODc × 100.
Chlorpyrifos is the positive control in acetone (“five concentration gradients”). The substrate is acetylcholine, not acetylthiocholine.
","page":4,"rects":[[50.72,29.11,92.21,67.13]]},{"id":"lc50-ladder","cat":"crit","crit":[5],"title":"LC50 ladder and endpoint","body":"Ladder 50, 100, 200, 400, 800 µg/mL; dead larvae counted at 72 h; corrected (Abbott-type) mortality; Excel + SPSS 20.0 probit regression gives the LC50. This belongs in each LC50 row's instructions, with no LC50 value leaking into them.
Units must be µg/mL: writing the ladder as 50–800 mg/mL is a heavy 1000× error.
","page":4,"rects":[[7.89,42.18,49.38,71.17]]},{"id":"reversibility","cat":"crit","crit":[5],"title":"Reversibility series (context for Ki)","body":"Enzyme at 0.125, 0.25, 0.5, 1.0 U/mL with inhibitor at 3.75–60 µg/mL, then 1.5 mM ACh + 0.3 mM DTNB. Worth carrying in the Ki instructions.
Figure 6A–B plots 0.2–0.8 U/mL instead; prefer the methods text, and flagging the mismatch is good.
","page":4,"rects":[[50.72,70.34,92.21,88.75]]},{"id":"lc90-symptoms","cat":"excl","crit":[],"title":"Symptoms and the LC90 dose","body":"The LC90 used here is never printed, so an invented LC90 is a hallucination. The Figure 4 symptom stages are qualitative: no rows.
","page":4,"rects":[[7.89,73.31,49.38,88.75]]},{"id":"ki-method","cat":"crit","crit":[4,5],"title":"Ki instructions","body":"Fixed enzyme; inhibitor at 3.75–60 µg/mL (100 µL, 37 °C, 1 h, acetone control); then 50 µL 0.3 mM DTNB + ACh at 0.325, 0.75, 1.5, 3, 6 mM (keep 0.325 as printed); OD at 412 nm after 60 s gives ΔA/min.
Lineweaver–Burk gives the inhibition type; the Dixon plot (Figure 6E/F, ACh 3 and 6 mM) gives Ki.
","page":5,"rects":[[7.89,9.15,49.38,40.87]]},{"id":"figure1","cat":"excl","crit":[1],"title":"Figure 1: bars without numbers","body":"The n-BuOH bars and the absolute EtOAc 24 h bar carry no printed values. Leaving them out is correct; eyeballed values (e.g. n-BuOH “~28%”) or an invented 0 are hallucinations. The Duncan letters may go in result_text.
","page":5,"rects":[[52.57,10.57,90.19,40.28]]},{"id":"antifeedant-values","cat":"rows","crit":[1],"title":"Antifeedant rows (3)","body":"| extract | time | % | Duncan |
|---|---|---|---|
| PE layer | 8 h | 61.14 | a |
| PE layer | 24 h | 48.56 | a |
| EtOAc layer | 8 h | 24.91 | b |
antifeedant_percentage, unit %, n_samples 3. Effector: a generic crude-extract ref per layer, never the solvent molecule. Either time framing is fine (8 / 24 h or 0–8 / 8–24 h).
The EtOAc +16.65 points from 8 h to 24 h must appear somewhere: in result_text, as its own change row, or as a 41.56% row marked as derived. Reporting 16.65 as the EtOAc 24 h rate is an error.
","page":5,"rects":[[36.13,86.02,41.24,87.45],[74.64,42.98,79.93,44.41],[54.05,47.43,59.09,48.86],[87,47.43,92.03,48.86]]},{"id":"table1","cat":"rows","crit":[8],"title":"Table 1 rows (11)","body":"| fraction | mortality % | Duncan |
|---|---|---|
| Lx.2 | 17.57 | g |
| Lx.3 | 10.37 | h |
| Lx.4 | 22.54 | f |
| Lx.5 | 32.11 | d |
| Lx.6 | 29.68 | e |
| Lx.7 | 36.28 | c |
| Lx.8 | 41.73 | b |
| Lx.9 | 53.76 | a |
| Lx.10 | 39.25 | b |
| Lx.11 | 27.14 | e |
| Lx.12 | 15.93 | g |
time_response, mortality_percentage %, diet film at 2 mg/mL. Evaluation time and replicates aren't stated, so n_samples is null (asserting 72 h or 2 is minor). Each fraction is a generic effector described as a silica-gel gradient fraction of the EtOAc layer.
Lx.1 is printed “–”: omit it with a note, or a 0 row that explains the dash. Lx.11 prints as “27.14.e”. Lx.8 is 41.73% although the text says the others “did not exceed 40%”. Errors: dose_response, an “ethyl acetate” effector, a nonzero Lx.1.
","page":6,"rects":[[7.89,9.15,47.87,40.28],[13.99,42.87,27.48,44.3]]},{"id":"compounds","cat":"crit","crit":[3,6],"title":"Compound identities","body":"1 N-p-coumaroyl tyramine, white non-crystalline powder, ESI-MS 283.3624 [M+]. 2 N-trans-feruloyl tyramine, 314.1387 [M+H]+.
Diagnostic NMR: trans olefin J 15.7–15.8 Hz; ethylene triplets δH 3.45 / 2.74; amide C=O δC 169.24 / 169.17; the OCH3 of 2 at δH 3.86, δC 55.36. Full peak lists are welcome but not required; transcribe as printed (1 lists 7.43 twice).
NMR shifts as physicochemical rows are minor over-extraction; tR and m/z rows are tolerated.
","page":6,"rects":[[7.89,51.56,49.38,70.1],[50.72,9.39,92.21,31.96]]},{"id":"route-screen","cat":"ctx","crit":[2],"title":"Which route is which","body":"§3.3 calls 47.97 µg/mL the contact LC50, which confirms Table 2's mapping: “gastrotoxic” = dietary, “poisoning” = contact. The 200 µg/mL single-dose screen (Figures 3A–B) is plot-only: no rows.
","page":6,"rects":[[50.72,34.34,92.21,70.81]]},{"id":"figure2","cat":"crit","crit":[6],"title":"Structures: what counts as a pass","body":"No SMILES, InChI or CAS is printed. A structure-denoting name in iupac_name (or name) plus a description of this drawing passes, and a null SMILES is correct.
A SMILES is fine only if it encodes the drawing exactly: trans C=C conjugated to the amide C=O, para-OH on both rings, a plain –CH2–CH2– linker, and for 2 a methoxy ortho to the cinnamoyl OH. Any wrong feature is heavy.
The figure labels 1 “N-p-Coumaroyloctopamine”, but the drawing has no side-chain OH and the text says tyramine. “Octopamine” is an error; flagging the conflict is good.
","page":6,"rects":[[9.57,73.54,86.5,90.77]]},{"id":"figure3","cat":"excl","crit":[],"title":"Figure 3: plot-only","body":"24/48/72 h mortality at 200 µg/mL (A–B) and mortality at each dose (C–D) have no printed numbers. Digitized values are hallucinations.
","page":7,"rects":[[9.57,10.57,87.84,49.9]]},{"id":"methoxy-c3","cat":"trap","crit":[3,6],"title":"Methoxy position","body":"The extra methoxy of 2 is at C-3′ (here, the NMR at δC 149.27, and Figure 2). §3.6 later says “C-6”; that is a misprint.
","page":7,"rects":[[8,57.83,19.97,59.33]]},{"id":"table2","cat":"rows","crit":[2],"title":"Table 2 rows (4 LC50s)","body":"| route | cpd | µg/mL | result_text |
|---|---|---|---|
| dietary | 1 | 305.25 | y = 46.962x − 66.683, R² 0.9888 |
| dietary | 2 | 89.88 | y = 44.673x − 37.275, R² 0.975 |
| contact | 1 | 92.96 | y = 45.083x − 38.738, R² 0.9972 |
| contact | 2 | 47.97 | y = 39.153x − 15.807, R² 0.9829 |
unit=\"ug/mL\"; mg/mL is a heavy 1000× error. “Gastrotoxic” = dietary, “poisoning” = contact. Equation and R² go in result_text, with no invented CI. n_samples 2 dietary, 3 contact. Target: organism S. frugiperda.
Check: setting y = 50 reproduces each LC50 (10^(116.683/46.962) ≈ 305.2).
","page":7,"rects":[[7.89,76.87,86.5,89.35]]},{"id":"figure4","cat":"excl","crit":[],"title":"Figure 4: qualitative","body":"Excitement, spasm, paralysis, death. No rows.
","page":8,"rects":[[9.57,10.57,86.5,42.89]]},{"id":"figure5","cat":"excl","crit":[4],"title":"Figure 5: plot-only","body":"5A shows in-vivo AChE activity (U/g) at LC50 and 2×LC50; 5B shows % inhibition per concentration. No printed numbers, so no rows; the IC50s come from Table 3.
","page":8,"rects":[[9.57,66.18,90.19,90.77]]},{"id":"table3","cat":"rows","crit":[4],"title":"Table 3 rows (3 IC50s)","body":"| cpd | µg/mL | ± | Duncan | |
|---|---|---|---|---|
| 1 | 23.34 | 2.72 | c | R² 0.9834 |
| 2 | 19.71 | 1.98 | b | R² 0.9692 |
| chlorpyrifos | 2.57 | 0.37 | a | R² 0.9763 |
Unit ug/mL. ± → uncertainty_value; the paper doesn't say SD or SE, so uncertainty_type is null (asserting SD is minor). n_samples null. The chlorpyrifos row is expected; omitting it is a minor miss.
The caption copies Table 2's (“two modes of action”), but this table is the in-vitro AChE IC50.
","page":9,"rects":[[7.89,9.15,44.51,21.27]]},{"id":"figure6","cat":"excl","crit":[4],"title":"Figure 6: kinetics plots only","body":"No Km, Vmax, rate or intercept is printed, so michaelis_menten stays empty and nothing here becomes a row. The 1/S axis reads µM⁻¹ although the ACh ladder is in mM; prefer the methods.
","page":9,"rects":[[9.57,24.59,87.17,87.09]]},{"id":"ic50-restated","cat":"ctx","crit":[4],"title":"IC50s restated","body":"The same three Table 3 values, not extra rows. Figure 5B is plot-only.
","page":10,"rects":[[7.89,9.39,49.38,36.48]]},{"id":"ki-values","cat":"rows","crit":[4],"title":"Ki rows (2)","body":"| cpd | Ki, µg/mL |
|---|---|
| 1 | 23.76 |
| 2 | 20.79 |
dose_response with value_type=\"Ki\", no uncertainty, never michaelis_menten or binding_affinity. result_text may say reversible and competitive (Km rises, Vmax unchanged); those stay qualitative.
Ki and IC50 nearly coincide (1: 23.76 vs 23.34; 2: 20.79 vs 19.71), so swapping metric or compound is an error. “C-6” in this paragraph is a misprint for C-3′.
","page":10,"rects":[[22.42,85.31,35.8,86.73],[80.94,15.5,84.64,16.93],[72.64,16.99,76.35,18.42],[78.64,22.93,92.02,24.36]]},{"id":"conclusion","cat":"ctx","crit":[],"title":"Conclusion restates","body":"LC50 47.97, IC50 19.71 ± 1.98 and the two Ki values again: the same rows, not new ones.
","page":10,"rects":[[50.72,29.7,92.21,63.21]]},{"id":"supplement","cat":"ctx","crit":[],"title":"source_flags","body":"The paper points to online Supplementary Material that isn't in the PDF, so has_missing_supplemental_files = true. No specific Table S or Figure S is cited, so false is only a minor miss.