Student or educator? Sign up with your school email and get 6 months of Pro free.

The DOI matches its Crossref record; 95 of the 99 references it cites were found in a registry or an index; 48 of the 101 sources its claims cite were checked for support.

HIF inhibitor 32-134D eradicates murine hepatocellular carcinoma in combination with anti-PD1 therapy

Open access
Salman, Shaima, Meyers, David J., Wicks, Elizabeth E., Lee, Sophia N., Datan, Emmanuel, Thomas, Aline M., Anders, Nicole M., Hwang, Yousang, Lyu, Yajing, Yang, Yongkang, Jackson, Walter, Dordai, Dominic, Rudek, Michelle A., Semenza, Gregg L.·2022·Journal of Clinical Investigation

Abstract

Hepatocellular carcinoma (HCC) is a major cause of cancer mortality worldwide and available therapies, including immunotherapies, are ineffective for many patients. HCC is characterized by intratumoral hypoxia, and increased expression of hypoxia-inducible factor 1α (HIF-1α) in diagnostic biopsies is associated with patient mortality...

Overview

CoChat Summary

## TL;DR

This paper identifies 32-134D, a low-molecular-weight HIF inhibitor, as a suppressor of HCC growth that blocks both HIF-1 and HIF-2 activity in human and mouse HCC models. In immunocompetent Hepa1-6 mice, adding 32-134D to anti-PD1 increased complete tumor eradication from 25% to 67%.

## Key findings

  • In hypoxic Hep3B cells, 32-134D significantly inhibited HIF-dependent reporter activity with an IC50 of 2.9 μM, and 252 analogs yielded 27 compounds with HIF-transcriptional IC50 values below 3.3 μM.
  • In Hep3B xenografts, daily intraperitoneal 32-134D caused partial growth inhibition at 20 mg/kg and maximal inhibition at 40 and 80 mg/kg; treated tumors showed decreased mass and reduced CD31-positive blood vessel area.
  • RNA-seq in hypoxic Hep3B cells showed that 3,326 of 3,849 hypoxia-induced genes (86%) were inhibited by 32-134D; overall, 4,689 genes were inhibited, and a 15-gene HIF signature was associated with survival of 364 HCC patients (82 months vs 42 months, P = 0.006).
  • In syngeneic Hepa1-6 tumors, 32-134D increased complete responses to anti-PD1 from 25% (3/12) to 67% (8/12), while 32-134D alone eradicated tumors in 4 of 12 mice.
  • 32-134D shifted the tumor immune microenvironment toward antitumor immunity: it increased CD8+IFN-γ+ effector T cells, activated CD8+ T cells, and activated NK cells, and decreased TAMs and M-MDSCs; the effector T-cell:TAM ratio increased 3-fold.
  • Safety signals were limited in mice: no changes were seen in appearance, behavior, body weight, hemoglobin, hematocrit, serum EPO, or histology of brain, colon, heart, kidney, liver, lungs, or small intestine after 14 days at 40 or 80 mg/kg/day.

## Methods

  • The study combined an NCI-60 virtual screen, chemical synthesis of lead analogs, in vitro reporter assays and RT-qPCR, mouse xenograft and syngeneic HCC models, flow cytometry, ELISA, immunohistochemistry, and RNA-seq.
  • Human Hep3B and mouse Hepa1-6 HCC cells were used; Hep3B xenografts were grown in nude mice, and Hepa1-6 tumors were grown in C57L mice, with treatment by daily intraperitoneal 32-134D, anti-PD1, or both.
  • Main analyses included FLuc/RLuc reporter assays under 1% O2, RT-qPCR/RNA-seq with differential expression thresholds of FDR < 0.05 and fold change > 1.5, tumor immune-cell profiling by flow cytometry, and pharmacokinetics by LC-MS/MS.

## Limitations

  • The paper states that further studies are required to determine whether steady-state plasma levels of 32-134D are achieved with daily administration.
  • The work is preclinical and limited to cell lines and mouse models; clinical efficacy in humans is not tested in the paper.
  • Evident from the text: some mechanistic claims about stromal versus tumor-cell effects are inferential, because differential effects in tumors versus cultured Hepa1-6 cells suggest but do not directly prove the responsible cell types.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1172/jci156774

Cites

95 of 95 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Identification of HIF inhibitors. (A) Hep3B-c1 cells were stably transfected with firefly luciferase (FLuc) reporter p2.1, which contains a hypoxia-response element (HRE), and Renilla luciferase (RLuc) reporter pSVR. (B) Hep3B-c1 cells were incubated with vehicle (Veh; 0.1% DMSO; blue bars) or 10 μM 32-134D (red bars) at 20% O2 (n = 6) or 1% O2 (n = 12) for 24 hours. Cell lysates were assayed for Fluc/Rluc activity (mean ± SEM); *P < 0.05 versus vehicle (χ2 test). (C) Chemical structures and IC50 values for HIF inhibitors. (D and E) Hep3B cells were exposed to 20% O2 with vehicle, or 1% O2 with vehicle (blue bar), 32-134D (red bars), 33-063 (green bars), or PT2385 (brown bars) for 24 hours and CA9 (D) and EPO (E) mRNAs were quantified by RT-qPCR. Data are presented as mean ± SEM (n = 3). *P < 0.05 versus 20% O2-vehicle; **P < 0.01 versus 1% O2-vehicle (ANOVA with Bonferroni’s post hoc test); NS, not significantly different from 1% O2-vehicle. (F) Hep3B cells were exposed to 20% or 1% O2 in the presence of vehicle or 1% O2 in the presence of 5 μM 32-134D (n = 3 each) for 24 hours. RNA sequencing identified genes with hypoxia-induced expression (blue circle) and genes that were inhibited by 32-134D (orange circle), based on FDR < 0.05 and fold change > 1.5. (G) Hep3B cells were exposed to 20% or 1% O2 for 24 hours in the presence of vehicle or 5 μM 32-134D, nuclear extracts were prepared, and immunoblot assays were performed. (H) Hep3B cells were exposed to 20% or 1% O2 for 24 hours with vehicle, 5 μM 32-134D, or 5 μM MG132 (during last 8 hours of exposure), nuclear extracts were prepared, and immunoblot assays were performed.
Figure 1 Identification of HIF inhibitors. (A) Hep3B-c1 cells were stably transfected with firefly luciferase (FLuc) reporter p2.1, which contains a hypoxia-response element (HRE), and Renilla luciferase (RLuc) reporter pSVR. (B) Hep3B-c1 cells were incubated with vehicle (Veh; 0.1% DMSO; blue bars) or 10 μM 32-134D (red bars) at 20% O2 (n = 6) or 1% O2 (n = 12) for 24 hours. Cell lysates were assayed for Fluc/Rluc activity (mean ± SEM); *P < 0.05 versus vehicle (χ2 test). (C) Chemical structures and IC50 values for HIF inhibitors. (D and E) Hep3B cells were exposed to 20% O2 with vehicle, or 1% O2 with vehicle (blue bar), 32-134D (red bars), 33-063 (green bars), or PT2385 (brown bars) for 24 hours and CA9 (D) and EPO (E) mRNAs were quantified by RT-qPCR. Data are presented as mean ± SEM (n = 3). *P < 0.05 versus 20% O2-vehicle; **P < 0.01 versus 1% O2-vehicle (ANOVA with Bonferroni’s post hoc test); NS, not significantly different from 1% O2-vehicle. (F) Hep3B cells were exposed to 20% or 1% O2 in the presence of vehicle or 1% O2 in the presence of 5 μM 32-134D (n = 3 each) for 24 hours. RNA sequencing identified genes with hypoxia-induced expression (blue circle) and genes that were inhibited by 32-134D (orange circle), based on FDR < 0.05 and fold change > 1.5. (G) Hep3B cells were exposed to 20% or 1% O2 for 24 hours in the presence of vehicle or 5 μM 32-134D, nuclear extracts were prepared, and immunoblot assays were performed. (H) Hep3B cells were exposed to 20% or 1% O2 for 24 hours with vehicle, 5 μM 32-134D, or 5 μM MG132 (during last 8 hours of exposure), nuclear extracts were prepared, and immunoblot assays were performed.CC BY · PubMed Central
Figure 2 Effect of 32-134D on Hep3B tumor xenograft growth and vascularization. (A) Female nude mice received a subcutaneous injection of 5 × 106 Hep3B cells. When tumors reached a volume of 150 mm3 (designated treatment day 1), the mice were randomized to receive a daily intraperitoneal injection of 32-134D at a dose of 0 (blue), 20 (black), 40 (red), or 80 (green) mg/kg. Data are presented as mean tumor volume (± SEM; n = 4 each). **P < 0.01, ***P < 0.001 (ANOVA with Bonferroni’s post hoc test). (B) Gross pathology of tumors harvested from vehicle-treated (top panel) and 32-134D–treated (bottom panel) mice. (C) Nuclear extracts prepared from tumors were assayed by immunoblotting using antibodies against the indicated proteins. (D) Total RNA was isolated from tumor tissue and analyzed by RT-qPCR using primers specific for the indicated mRNAs and results (mean ± SEM, n = 4) were normalized to the mean value for tumors from vehicle-treated mice. (E) ELISA for the indicated proteins was performed using aliquots of tumor lysates (mean ± SEM; n = 3–4 tumors each). *P < 0.05 (Mann-Whitney test). (F) Formalin-fixed and paraffin-embedded tumor sections were analyzed by immunohistochemistry using an antibody against CD31 to identify vascular endothelial cells. Scale bar: 100 μm. The total CD31+ vessel area per field was quantified using ImageJ (mean ± SEM; n = 4 tumors with 5 sections per tumor). *P < 0.05 (Student’s t test).
Figure 2 Effect of 32-134D on Hep3B tumor xenograft growth and vascularization. (A) Female nude mice received a subcutaneous injection of 5 × 106 Hep3B cells. When tumors reached a volume of 150 mm3 (designated treatment day 1), the mice were randomized to receive a daily intraperitoneal injection of 32-134D at a dose of 0 (blue), 20 (black), 40 (red), or 80 (green) mg/kg. Data are presented as mean tumor volume (± SEM; n = 4 each). **P < 0.01, ***P < 0.001 (ANOVA with Bonferroni’s post hoc test). (B) Gross pathology of tumors harvested from vehicle-treated (top panel) and 32-134D–treated (bottom panel) mice. (C) Nuclear extracts prepared from tumors were assayed by immunoblotting using antibodies against the indicated proteins. (D) Total RNA was isolated from tumor tissue and analyzed by RT-qPCR using primers specific for the indicated mRNAs and results (mean ± SEM, n = 4) were normalized to the mean value for tumors from vehicle-treated mice. (E) ELISA for the indicated proteins was performed using aliquots of tumor lysates (mean ± SEM; n = 3–4 tumors each). *P < 0.05 (Mann-Whitney test). (F) Formalin-fixed and paraffin-embedded tumor sections were analyzed by immunohistochemistry using an antibody against CD31 to identify vascular endothelial cells. Scale bar: 100 μm. The total CD31+ vessel area per field was quantified using ImageJ (mean ± SEM; n = 4 tumors with 5 sections per tumor). *P < 0.05 (Student’s t test).CC BY · PubMed Central
Figure 3 Effect of 32-134D treatment on hypoxia-induced gene expression in Hepa1-6 cells. (A–C) Cells were exposed to 20% O2 and vehicle (white bars), 1% O2 and vehicle (blue bars), or 1% O2 and 32-134D (red bars) for 24 hours and mRNAs were quantified by RT-qPCR and normalized to white (mean ± SEM, n = 4). *P < 0.05 versus white; #P < 0.05 versus blue (ANOVA with Bonferroni’s post hoc test).
Figure 3 Effect of 32-134D treatment on hypoxia-induced gene expression in Hepa1-6 cells. (A–C) Cells were exposed to 20% O2 and vehicle (white bars), 1% O2 and vehicle (blue bars), or 1% O2 and 32-134D (red bars) for 24 hours and mRNAs were quantified by RT-qPCR and normalized to white (mean ± SEM, n = 4). *P < 0.05 versus white; #P < 0.05 versus blue (ANOVA with Bonferroni’s post hoc test).CC BY · PubMed Central
Figure 4 Effect of anti-PD1 and 32-134D on Hepa1-6 tumor growth in syngeneic mice. C57L mice were injected with Hepa1-6 HCC cells subcutaneously and when tumors became palpable, they were randomized to receive intraperitoneal injection of vehicle (A) or 32-134D (40 mg/kg; B) daily; IgG2a isotype control (C) or anti-PD1 (D) antibody every 3 days; or both anti-PD1 and 32-134D (E). The percentage of mice in each treatment group with tumor eradication on day 34 is shown (F; blue, green, and red bars).
Figure 4 Effect of anti-PD1 and 32-134D on Hepa1-6 tumor growth in syngeneic mice. C57L mice were injected with Hepa1-6 HCC cells subcutaneously and when tumors became palpable, they were randomized to receive intraperitoneal injection of vehicle (A) or 32-134D (40 mg/kg; B) daily; IgG2a isotype control (C) or anti-PD1 (D) antibody every 3 days; or both anti-PD1 and 32-134D (E). The percentage of mice in each treatment group with tumor eradication on day 34 is shown (F; blue, green, and red bars).CC BY · PubMed Central
Figure 5 Effect of 32-134D on the tumor immune microenvironment. (A–H) C57L mice were injected with Hepa1-6 HCC cells subcutaneously and when tumors reached a volume of 200 mm3, the mice were treated with vehicle or 32-134D (40 mg/kg) by daily intraperitoneal injection for 8 days. Single-cell suspensions prepared from each tumor were analyzed by flow cytometry using fluorescent antibodies against the indicated cell surface proteins. The percentage of cells positive for the indicated markers is shown (mean ± SEM, n = 6). *P < 0.05 (Mann-Whitney test). M-MDSCs and G-MDSCs, monocytic and granulocytic myeloid-derived suppressor cells; TAMs, tumor-associated macrophages.
Figure 5 Effect of 32-134D on the tumor immune microenvironment. (A–H) C57L mice were injected with Hepa1-6 HCC cells subcutaneously and when tumors reached a volume of 200 mm3, the mice were treated with vehicle or 32-134D (40 mg/kg) by daily intraperitoneal injection for 8 days. Single-cell suspensions prepared from each tumor were analyzed by flow cytometry using fluorescent antibodies against the indicated cell surface proteins. The percentage of cells positive for the indicated markers is shown (mean ± SEM, n = 6). *P < 0.05 (Mann-Whitney test). M-MDSCs and G-MDSCs, monocytic and granulocytic myeloid-derived suppressor cells; TAMs, tumor-associated macrophages.CC BY · PubMed Central
Figure 6 Effect of 32-134D on intratumoral gene expression. (A–D) C57L mice were injected with Hepa1-6 HCC cells subcutaneously and when tumors reached a volume of 200 mm3, the mice were treated with vehicle (blue bars) or 32-134D (40 mg/kg; red bars) by daily intraperitoneal injection for 8 days. The tumors were harvested and mRNA was quantified by RT-qPCR and normalized to blue (mean ± SEM, n = 4). *P < 0.05 versus blue (ANOVA with Bonferroni’s post hoc test). (E) Effect of 32-134D on intratumoral expression of cytokines and chemokines. Total RNA isolated from tumors of 32-134D–treated versus vehicle-treated mice (n = 3 each) was analyzed using an RT-qPCR array and the ratio of mean expression (32-134D/vehicle) was determined. mRNAs with a significant difference between groups (P < 0.05, Student’s t test) are annotated.
Figure 6 Effect of 32-134D on intratumoral gene expression. (A–D) C57L mice were injected with Hepa1-6 HCC cells subcutaneously and when tumors reached a volume of 200 mm3, the mice were treated with vehicle (blue bars) or 32-134D (40 mg/kg; red bars) by daily intraperitoneal injection for 8 days. The tumors were harvested and mRNA was quantified by RT-qPCR and normalized to blue (mean ± SEM, n = 4). *P < 0.05 versus blue (ANOVA with Bonferroni’s post hoc test). (E) Effect of 32-134D on intratumoral expression of cytokines and chemokines. Total RNA isolated from tumors of 32-134D–treated versus vehicle-treated mice (n = 3 each) was analyzed using an RT-qPCR array and the ratio of mean expression (32-134D/vehicle) was determined. mRNAs with a significant difference between groups (P < 0.05, Student’s t test) are annotated.CC BY · PubMed Central
Figure 7 Effect of 32-134D on intratumoral expression of immunoregulatory proteins. (A–H) Lysates prepared from tumors of 32-134D–treated versus vehicle-treated mice were subjected to ELISA for secreted proteins mediating antitumor immunity (CXCL2, CXCL9, CXCL10) or immunosuppression (CXCL1, IL-6, IL-10, VEGFA). The data are presented as mean ± SEM (n = 4 each). *P < 0.05 (Mann-Whitney test).
Figure 7 Effect of 32-134D on intratumoral expression of immunoregulatory proteins. (A–H) Lysates prepared from tumors of 32-134D–treated versus vehicle-treated mice were subjected to ELISA for secreted proteins mediating antitumor immunity (CXCL2, CXCL9, CXCL10) or immunosuppression (CXCL1, IL-6, IL-10, VEGFA). The data are presented as mean ± SEM (n = 4 each). *P < 0.05 (Mann-Whitney test).CC BY · PubMed Central
Figure 8 Effect of 32-134D on red blood cell indices. Mice (n = 4 per group) were treated with vehicle (V) or 32-134D (40 mg/kg/day) for 14 days and peripheral blood was analyzed for red blood cell count (RBC), hemoglobin (Hgb), hematocrit (Hct), reticulocytes, mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), mean corpuscular volume (MCV), and absolute reticulocyte count (A) or EPO levels in serum (B).
Figure 8 Effect of 32-134D on red blood cell indices. Mice (n = 4 per group) were treated with vehicle (V) or 32-134D (40 mg/kg/day) for 14 days and peripheral blood was analyzed for red blood cell count (RBC), hemoglobin (Hgb), hematocrit (Hct), reticulocytes, mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), mean corpuscular volume (MCV), and absolute reticulocyte count (A) or EPO levels in serum (B).CC BY · PubMed Central
Figure 9 Concentration-time profiles of 32-134D in mice (n = 3 per time point) treated with a single dose of 32-134D. Plasma was obtained over 24 hours, with 32-134D concentrations determined by LC-MS/MS. Dashed line represents the in vitro IC50 of 32-134D (2.5 μM). Data points and error bars represent mean and SD of 3 replicates, respectively.
Figure 9 Concentration-time profiles of 32-134D in mice (n = 3 per time point) treated with a single dose of 32-134D. Plasma was obtained over 24 hours, with 32-134D concentrations determined by LC-MS/MS. Dashed line represents the in vitro IC50 of 32-134D (2.5 μM). Data points and error bars represent mean and SD of 3 replicates, respectively.CC BY · PubMed Central
Figure 10 Immunological and other effects of HIF inhibition by 32-134D on HCC progression. The figure summarizes the major HIF target genes that were analyzed in this study. It does not include the many HIF target genes that are involved in other critical aspects of HCC progression that were not analyzed in this study. CTL, cytotoxic T lymphocyte; NKC, natural killer cell.
Figure 10 Immunological and other effects of HIF inhibition by 32-134D on HCC progression. The figure summarizes the major HIF target genes that were analyzed in this study. It does not include the many HIF target genes that are involved in other critical aspects of HCC progression that were not analyzed in this study. CTL, cytotoxic T lymphocyte; NKC, natural killer cell.CC BY · PubMed Central

Verification

Claims the paper makes

24 Valid · 3 Not valid · 18 Not checked · 97 Own results and interpretations

85 more claims · 4 with their cited sources checked

Create a free account to see all 142 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 33
  • Probably not a substantial claim. 26
  • Borderline: one of the tests for a substantial claim is unclear. 1
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

7 more claims about this paper

Create a free account to see every claim other papers make about this one.

See all claims — free account

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

1 of 1

Cited works that were retracted

Works this paper cites that were later retracted, from the retraction notices publishers deposit. Major when the notice came out before this paper was published.

1 of 1

Citations

Cites

1 retracted

Show all 99 references

Cited by

See all 111

Metadata

Authors
Salman, S., Meyers, D. J., Wicks, E. E., Lee, S. N., Datan, E., Thomas, A. M., Anders, N. M., Hwang, Y., Lyu, Y., Yang, Y., Jackson, W., Dordai, D., Rudek, M. A., & Semenza, G. L.
Affiliations
  • Sidney Kimmel Cancer Center (US)
  • Johns Hopkins University (US)
Corresponding author
Semenza, Gregg L.
Venue
Journal of Clinical Investigation
Published
2022-05-02
Volume
132
Issue
9
ISSN
  • 0021-9738 (ISSN-L)
  • 1558-8238
OpenAlex
W4225257247
Cited by
111
References
99
Citations by year
2026: 20 · 2025: 34 · 2024: 18 · 2023: 29 · 2022: 10
FWCI
9.79
Type
journal article
Language
en
Open access
Locations
  • Journal of Clinical Investigation (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
  • American Cancer Society (RP-16-239-06-COUN)
  • Armstrong Family Foundation (n/a)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cancer Research › Cancer, Hypoxia, and Metabolism
  • Health Sciences › Medicine › Hepatology › Hepatocellular Carcinoma Treatment and Prognosis
  • Health Sciences › Medicine › Oncology › Cancer Immunotherapy and Biomarkers
Keywords
hepatocellular carcinoma, HIF-1α inhibition, HIF-2α inhibition, anti-PD-1 therapy, combination immunotherapy, tumor hypoxia, tumor immune microenvironment, tumor-associated macrophages, myeloid-derived suppressor cells, CD8+ T cells, natural killer cells, angiogenesis, glycolytic metabolism, immune evasion, antitumor immunity
MeSH terms
Animals, Hypoxia, Carcinoma, Hepatocellular, Humans, Liver Neoplasms, Neovascularization, Pathologic, Mice, Tumor Microenvironment

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: HIF inhibitor 32-134D eradicates murine hepatocellular carcinoma in combination with anti-PD1 therapy. https://app.cochat.ai/papers/10.1172/jci156774
The paper
Salman, S., Meyers, D. J., Wicks, E. E., Lee, S. N., Datan, E., Thomas, A. M., Anders, N. M., Hwang, Y., Lyu, Y., Yang, Y., Jackson, W., Dordai, D., Rudek, M. A., & Semenza, G. L. (2022). HIF inhibitor 32-134D eradicates murine hepatocellular carcinoma in combination with anti-PD1 therapy. Journal of Clinical Investigation. https://doi.org/10.1172/jci156774

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.