Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat
Sign Up
Cited by 5Malaria Research and ControlToxoplasma gondii Research StudiesBird parasitology and diseases

The DOI matches its Crossref record; 55 of the 55 references it cites were found in a registry or an index.

SIP2 is the master transcription factor of Plasmodium merozoite formation

Open access
Nishi, Tsubasa, Kaneko, Izumi, Yuda, Masao·2025·Science Advances
DOI 10.1126/sciadv.ads5458PMC PMC11927625

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

Malaria, one of the most serious infectious diseases worldwide, is caused by the proliferation of Plasmodium parasites through repeated cycles of intraerythrocytic development. The parasite replicates via schizogony in host erythrocytes, producing multiple progeny merozoites that invade new erythrocytes. Although merozoite formation is the most crucial step in malaria pathogenesis, its molecular mechanism remains unclear. SIP2 is an AP2 transcription factor expressed during schizogony and is particularly conserved among erythrocyte-infecting apicomplexan parasites. Here, we reveal that SIP2 in Plasmodium berghei (PbSIP2) functions as the master transcription factor for merozoite formation. Conditional disruption of pbsip2 resulted in developmental arrest before merozoite formation and notable down-regulation of merozoite-related genes. ChIP-seq showed that PbSIP2 comprehensively activated merozoite-related genes by binding to previously reported cis-regulatory elements of merozoite invasion-related genes, including the bipartite motif (TGCAN 4–6 GTGCA). Collectively, our results indicate that SIP2 is a transcription factor that establishes erythrocyte infectivity and may have an evolutionary origin from the common ancestor of erythrocyte-infecting apicomplexan parasites.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1126/sciadv.ads5458

Cites

55 of 55 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1. Sequence features of PbSIP2 and its expression during asexual-blood stage development. (A) Schematic illustration of PbSIP2. Blue boxes show the AP2 domains. The NLS was predicted using cNLS Mapper (http://nls-mapper.iab.keio.ac.jp/cgi-bin/NLS_Mapper_form.cgi) and indicated using a yellow box. (B) Alignment of amino acid (AA) sequences of PbSIP2 orthologs in Plasmodium by the ClustalW program in Mega X (P. berghei, PbANKA_0102900; P. falciparum, PF3D7_0604100; Plasmodium malariae, PmUG01_11058700; Plasmodium vivax, PVP01_1144800; Plasmodium knowlesi, PKNH_1146400; Plasmodium gallinaceum, PGAL8A_00132600). Amino acid sequences conserved in all aligned sequences are indicated by blue boxes. Yellow and red bars indicate a β sheet and α helix, respectively. Blue bars show a sequence for an AP2 domain. (C) Alignment of amino acid sequences of PbSIP2, PfSIP2, and BLASTP-detected proteins (Hepatocystis spp., VWU52144; Babesia Bovis, XP_001610230; Theileria Parva, XP_764727). Amino acid sequences conserved in all aligned sequences are indicated by yellow boxes. Regions wherein more than 10 amino acids are continuously conserved are indicated with asterisks. (D) Taxonomic tree of Apicomplexa. Orders in which PbSIP2 is conserved are indicated in a blue box. (E) Schematic illustration of putative SIP2 orthologs in Apicomplexa. Blue boxes show the AP2 domains, and NLSs are indicated using a yellow box. (F) Expression of PbSIP2 in asexual blood stage of the PbSIP2::GFP. Rings and trophozoites were detected from the peripheral blood, and schizonts were observed in synchronized cultures. Nuclei were stained with Hoechst 33342. Scale bar, 5 μm. (G) Quantification of GFP-positive cells for PbSIP2::GFP with each number of nuclei. Cells were counted by fluorescent microscopic analysis. At least 30 cells were assessed for schizonts with each number of nuclei [>8 (mero.) indicates merozoites]. hpi, hours post-injection.
Fig. 1. Sequence features of PbSIP2 and its expression during asexual-blood stage development. (A) Schematic illustration of PbSIP2. Blue boxes show the AP2 domains. The NLS was predicted using cNLS Mapper (http://nls-mapper.iab.keio.ac.jp/cgi-bin/NLS_Mapper_form.cgi) and indicated using a yellow box. (B) Alignment of amino acid (AA) sequences of PbSIP2 orthologs in Plasmodium by the ClustalW program in Mega X (P. berghei, PbANKA_0102900; P. falciparum, PF3D7_0604100; Plasmodium malariae, PmUG01_11058700; Plasmodium vivax, PVP01_1144800; Plasmodium knowlesi, PKNH_1146400; Plasmodium gallinaceum, PGAL8A_00132600). Amino acid sequences conserved in all aligned sequences are indicated by blue boxes. Yellow and red bars indicate a β sheet and α helix, respectively. Blue bars show a sequence for an AP2 domain. (C) Alignment of amino acid sequences of PbSIP2, PfSIP2, and BLASTP-detected proteins (Hepatocystis spp., VWU52144; Babesia Bovis, XP_001610230; Theileria Parva, XP_764727). Amino acid sequences conserved in all aligned sequences are indicated by yellow boxes. Regions wherein more than 10 amino acids are continuously conserved are indicated with asterisks. (D) Taxonomic tree of Apicomplexa. Orders in which PbSIP2 is conserved are indicated in a blue box. (E) Schematic illustration of putative SIP2 orthologs in Apicomplexa. Blue boxes show the AP2 domains, and NLSs are indicated using a yellow box. (F) Expression of PbSIP2 in asexual blood stage of the PbSIP2::GFP. Rings and trophozoites were detected from the peripheral blood, and schizonts were observed in synchronized cultures. Nuclei were stained with Hoechst 33342. Scale bar, 5 μm. (G) Quantification of GFP-positive cells for PbSIP2::GFP with each number of nuclei. Cells were counted by fluorescent microscopic analysis. At least 30 cells were assessed for schizonts with each number of nuclei [>8 (mero.) indicates merozoites]. hpi, hours post-injection.CC BY · PubMed Central
Fig. 2. Conditional knockout of pbsip2 by DiCre-mediated recombination. (A) Schematic illustration of DiCre-mediated recombination at the pbsip2 locus of pbsip2-cKO. Recombination occurs between two loxP sites to excise pbsip2 from the genome in the presence of rapamycin. P1 to P3 indicate primers used for genotyping PCR. (B) Outline of conditional knockout experiments using pbsip2-cKO. First, the whole blood from mice infected with pbsip2-cKO was cultured. After 16 hours of culture, schizonts were harvested by density gradient centrifugation and injected into mice. At 6 hpi, the whole blood was harvested and split into two cultures: one with 10 nM rapamycin and the other with DMSO. (C) A gel image of the genotyping PCR analysis for pbsip2-cKORapa− and pbsip2-cKORapa+ performed at 16 hpi (a representative image from the first of three replicates). Primers used are illustrated in (A). (D) Representative Giemsa-stained images of pbsip2-cKORapa− and pbsip2-cKORapa+ at 26 hpi. Mature schizonts with merozoites inside are indicated by arrows. Scale bar, 5 μm. (E) Ratio of schizonts with different numbers of nuclei in pbsip2-cKORapa− and pbsip2-cKORapa+ at 26 hpi (left) and 28 hpi (right). Error bars indicate the SEM value from three biologically independent experiments. The P values were calculated by two-tailed Student’s t test (n.s. stands for not significant, P value > 0.05). (F) Ratio of schizonts with different numbers of Hoechst foci in pbsip2-cKORapa− and pbsip2-cKORapa+ at 28 hpi. (G) Flow cytometric analysis of pbsip2-cKORapa− and pbsip2-cKORapa+ at 28 hpi. Mean Hoechst intensities of schizont populations are shown. Error bars indicate the SE from three independent experiments. The P value was calculated using a two-tailed Student’s t test. (H) Model of the P. berghei schizont development in synchronized cultures. The period for PbSIP2 expression is indicated as a green box. hpi, hours post-injection.
Fig. 2. Conditional knockout of pbsip2 by DiCre-mediated recombination. (A) Schematic illustration of DiCre-mediated recombination at the pbsip2 locus of pbsip2-cKO. Recombination occurs between two loxP sites to excise pbsip2 from the genome in the presence of rapamycin. P1 to P3 indicate primers used for genotyping PCR. (B) Outline of conditional knockout experiments using pbsip2-cKO. First, the whole blood from mice infected with pbsip2-cKO was cultured. After 16 hours of culture, schizonts were harvested by density gradient centrifugation and injected into mice. At 6 hpi, the whole blood was harvested and split into two cultures: one with 10 nM rapamycin and the other with DMSO. (C) A gel image of the genotyping PCR analysis for pbsip2-cKORapa− and pbsip2-cKORapa+ performed at 16 hpi (a representative image from the first of three replicates). Primers used are illustrated in (A). (D) Representative Giemsa-stained images of pbsip2-cKORapa− and pbsip2-cKORapa+ at 26 hpi. Mature schizonts with merozoites inside are indicated by arrows. Scale bar, 5 μm. (E) Ratio of schizonts with different numbers of nuclei in pbsip2-cKORapa− and pbsip2-cKORapa+ at 26 hpi (left) and 28 hpi (right). Error bars indicate the SEM value from three biologically independent experiments. The P values were calculated by two-tailed Student’s t test (n.s. stands for not significant, P value > 0.05). (F) Ratio of schizonts with different numbers of Hoechst foci in pbsip2-cKORapa− and pbsip2-cKORapa+ at 28 hpi. (G) Flow cytometric analysis of pbsip2-cKORapa− and pbsip2-cKORapa+ at 28 hpi. Mean Hoechst intensities of schizont populations are shown. Error bars indicate the SE from three independent experiments. The P value was calculated using a two-tailed Student’s t test. (H) Model of the P. berghei schizont development in synchronized cultures. The period for PbSIP2 expression is indicated as a green box. hpi, hours post-injection.CC BY · PubMed Central
Fig. 3. ChIP-seq analysis of PbSIP2. (A) Integrative Genomics Viewer (IGV) images for ChIP-seq experiments 1 and 2 of PbSIP2 on the chromosome 1. Histograms show row read coverage of ChIP data at each base. Scales are indicated in square brackets. (B) A Venn diagram showing the number of overlapping peaks between experiment 1 and experiment 2. (C) Motif enriched within 50 bp from ChIP-seq peak summits of PbSIP2. The logo was depicted using WebLogo (https://weblogo.berkeley.edu/logo.cgi). P valuelimit is a P value less than 5.0 × 10−324 (the smallest positive real number on the R platform). (D) Distance between peak summits identified in the ChIP-seq and the nearest GTGCA motifs. (E) Classification of functionally annotated target genes of PbSIP2 into 16 characteristic groups. The number of genes categorized into each group is indicated in the graph. (F) IGV images showing ChIP-seq peaks upstream of the PbSIP2 target genes. Histograms show row read coverage of ChIP data at each base. Locations of GTGCA motif are indicated as a blue bar. The scales are shown in square brackets. (G) Schematic illustration of P. berghei merozoite. Characteristic structures of the merozoite and PbSIP2 targets that belong to each structure are shown.
Fig. 3. ChIP-seq analysis of PbSIP2. (A) Integrative Genomics Viewer (IGV) images for ChIP-seq experiments 1 and 2 of PbSIP2 on the chromosome 1. Histograms show row read coverage of ChIP data at each base. Scales are indicated in square brackets. (B) A Venn diagram showing the number of overlapping peaks between experiment 1 and experiment 2. (C) Motif enriched within 50 bp from ChIP-seq peak summits of PbSIP2. The logo was depicted using WebLogo (https://weblogo.berkeley.edu/logo.cgi). P valuelimit is a P value less than 5.0 × 10−324 (the smallest positive real number on the R platform). (D) Distance between peak summits identified in the ChIP-seq and the nearest GTGCA motifs. (E) Classification of functionally annotated target genes of PbSIP2 into 16 characteristic groups. The number of genes categorized into each group is indicated in the graph. (F) IGV images showing ChIP-seq peaks upstream of the PbSIP2 target genes. Histograms show row read coverage of ChIP data at each base. Locations of GTGCA motif are indicated as a blue bar. The scales are shown in square brackets. (G) Schematic illustration of P. berghei merozoite. Characteristic structures of the merozoite and PbSIP2 targets that belong to each structure are shown.CC BY · PubMed Central
Fig. 4. Differential expression analysis between pbsip2-cKORapa− and pbsip2-cKORapa+ by RNA-seq. (A) Volcano plot showing differential expression of genes between pbsip2-cKORapa− and pbsip2-cKORapa+. Blue dots represent target genes of PbSIP2, and a red dot represents pbsip2. A horizontal line indicates a P value of 0.05 and two vertical lines indicate a log2(fold change) of 1 and −1. (B) A Venn diagram showing an overlap between the genes significantly downregulated in pbsip2-cKORapa+ and the PbSIP2 target genes. (C) Six-base motifs enriched in the upstream (300 to 1200 bp from start codons) of genes down-regulated in pbsip2-cKORapa+. Motifs including GTGCA are indicated in light yellow. (D) Classification of the downregulated genes into characteristic groups. (E) Section of the volcano plot from (A). Genes significantly down-regulated in pbsip2-cKORapa+ compared with pbsip2-cKORapa− are shown. Green dots represent a rhoptry gene, and names of some rhoptry genes are shown.
Fig. 4. Differential expression analysis between pbsip2-cKORapa− and pbsip2-cKORapa+ by RNA-seq. (A) Volcano plot showing differential expression of genes between pbsip2-cKORapa− and pbsip2-cKORapa+. Blue dots represent target genes of PbSIP2, and a red dot represents pbsip2. A horizontal line indicates a P value of 0.05 and two vertical lines indicate a log2(fold change) of 1 and −1. (B) A Venn diagram showing an overlap between the genes significantly downregulated in pbsip2-cKORapa+ and the PbSIP2 target genes. (C) Six-base motifs enriched in the upstream (300 to 1200 bp from start codons) of genes down-regulated in pbsip2-cKORapa+. Motifs including GTGCA are indicated in light yellow. (D) Classification of the downregulated genes into characteristic groups. (E) Section of the volcano plot from (A). Genes significantly down-regulated in pbsip2-cKORapa+ compared with pbsip2-cKORapa− are shown. Green dots represent a rhoptry gene, and names of some rhoptry genes are shown.CC BY · PubMed Central
Fig. 5. DNA binding property of PbSIP2 and function of its binding motif as a cis-regulatory element. (A) IGV image for DIP-seq using recombinant AP2 domains of PbSIP2 on the chromosome 1. Histograms show row read coverage of IP data at each base. (B) A motif enriched within 50 bp from peak summits of DIP-seq. The logo was depicted using WebLogo. P valuelimit is P value less than 5.0 × 10−324 (smallest positive real number on the R platform). An arrow indicates slight enrichment of G and A in the logo. (C) Six-base motifs enriched in the DIP-seq peaks. RGTGCA motif is indicated in yellow, and its one-base shifted motifs are indicated in light yellow. P valueave is the average P value derived from the seven separate motif enrichment analyses (the detail is described in Materials and Methods). Motifs were ranked along their P valueave. (D) Violin plot showing the distribution of fold enrichment values for the ChIP-seq peaks with RGTGCA (blue) and others (gray). The corresponding box plots are shown inside. The P value was calculated by two-tailed Student’s t test. (E) IGV image showing the ChIP-seq peak upstream of rop14 and genomic sequence around RGTGCA in the peak region. Sanger sequence result for PbSIP2::GFPCas9_cismut is shown under the wild-type sequence. The target region of single guide RNA used for generating PbSIP2::GFPCas9_cismut is indicated on the sequence. (F) ChIP-qPCR analysis of PbSIP2 at the mutated site upstream of rop14. Error bars indicate the standard error from three biologically independent experiments. The P values were calculated by two-tailed Student’s t test (i.e., P value > 0.01) [rop14: PBANKA_0111600, ralp1: PBANKA_0619700, ron4: PBANKA_0932000, trap: PBANKA_1349800]. (G) RT-qPCR analysis of rop14 transcripts between PbSIP2::GFPCas9 and PbSIP2::GFPCas9_cismut. Error bars and P value were generated same as (F).
Fig. 5. DNA binding property of PbSIP2 and function of its binding motif as a cis-regulatory element. (A) IGV image for DIP-seq using recombinant AP2 domains of PbSIP2 on the chromosome 1. Histograms show row read coverage of IP data at each base. (B) A motif enriched within 50 bp from peak summits of DIP-seq. The logo was depicted using WebLogo. P valuelimit is P value less than 5.0 × 10−324 (smallest positive real number on the R platform). An arrow indicates slight enrichment of G and A in the logo. (C) Six-base motifs enriched in the DIP-seq peaks. RGTGCA motif is indicated in yellow, and its one-base shifted motifs are indicated in light yellow. P valueave is the average P value derived from the seven separate motif enrichment analyses (the detail is described in Materials and Methods). Motifs were ranked along their P valueave. (D) Violin plot showing the distribution of fold enrichment values for the ChIP-seq peaks with RGTGCA (blue) and others (gray). The corresponding box plots are shown inside. The P value was calculated by two-tailed Student’s t test. (E) IGV image showing the ChIP-seq peak upstream of rop14 and genomic sequence around RGTGCA in the peak region. Sanger sequence result for PbSIP2::GFPCas9_cismut is shown under the wild-type sequence. The target region of single guide RNA used for generating PbSIP2::GFPCas9_cismut is indicated on the sequence. (F) ChIP-qPCR analysis of PbSIP2 at the mutated site upstream of rop14. Error bars indicate the standard error from three biologically independent experiments. The P values were calculated by two-tailed Student’s t test (i.e., P value > 0.01) [rop14: PBANKA_0111600, ralp1: PBANKA_0619700, ron4: PBANKA_0932000, trap: PBANKA_1349800]. (G) RT-qPCR analysis of rop14 transcripts between PbSIP2::GFPCas9 and PbSIP2::GFPCas9_cismut. Error bars and P value were generated same as (F).CC BY · PubMed Central
Fig. 6. Association of PbSIP2 with SBM. (A) GTGCA-containing bipartite motifs detected in PbSIP2 peaks upstream of rhoptry genes. GTGCA and GTGCA-like motifs are indicated by blue and gray, respectively. (B) 3D structure of the PbSIP2 AP2 domains (positions 87 to 214) predicted using AlphaFold2 (35). On the right, a B-DNA model depicted using GraphiteLifeExplorer (50) is shown. (C) Four–base pair motifs enriched within 5′-side 10 bp of GTGCA in the ChIP-seq peak regions. The ChIP-seq data presented in Fig. 3 were used for the analysis. The P values were calculated by Fisher’s exact test. TGCA motif and its one base–shifted motifs are indicated by blue. (D) Number of ChIP-seq peaks containing SBM-Nx with different length of interval sequence in the bipartite motif. (E) Pie graph showing the number of ChIP-seq peaks with fSBM (full SBM) and mSBM (SBM with a single mismatch in the 5′ TGCA). (F) Violin plot showing distribution of fold enrichment values for the ChIP-seq peaks with fSBM, mSBM, and others (left) and violin plot for those with SBM and RGTGCA (right). The corresponding box plots are shown inside. The P values were calculated by two-tailed Student’s t test. (G) Pie graph showing the number of DIP-seq peaks containing fSBM and mSBM. The DIP-seq data presented in Fig. 5 were used for the analysis. (H) Violin plot showing distribution of fold enrichment values for the DIP-seq peaks with fSBM, mSBM, and others. (I) IGV images showing ChIP-seq peaks on the upstream of PbSIP2 target genes. Red and pink bars indicate location of the fSBM and mSBM, respectively. Scales are indicated in square brackets. (J) Histogram showing the number of genes against the log2(fold change) value in the differential expression analysis between pbsip2-cKORapa− and pbsip2-cKORapa+. The percentages of genes associated with SBM are indicated on the graph.
Fig. 6. Association of PbSIP2 with SBM. (A) GTGCA-containing bipartite motifs detected in PbSIP2 peaks upstream of rhoptry genes. GTGCA and GTGCA-like motifs are indicated by blue and gray, respectively. (B) 3D structure of the PbSIP2 AP2 domains (positions 87 to 214) predicted using AlphaFold2 (35). On the right, a B-DNA model depicted using GraphiteLifeExplorer (50) is shown. (C) Four–base pair motifs enriched within 5′-side 10 bp of GTGCA in the ChIP-seq peak regions. The ChIP-seq data presented in Fig. 3 were used for the analysis. The P values were calculated by Fisher’s exact test. TGCA motif and its one base–shifted motifs are indicated by blue. (D) Number of ChIP-seq peaks containing SBM-Nx with different length of interval sequence in the bipartite motif. (E) Pie graph showing the number of ChIP-seq peaks with fSBM (full SBM) and mSBM (SBM with a single mismatch in the 5′ TGCA). (F) Violin plot showing distribution of fold enrichment values for the ChIP-seq peaks with fSBM, mSBM, and others (left) and violin plot for those with SBM and RGTGCA (right). The corresponding box plots are shown inside. The P values were calculated by two-tailed Student’s t test. (G) Pie graph showing the number of DIP-seq peaks containing fSBM and mSBM. The DIP-seq data presented in Fig. 5 were used for the analysis. (H) Violin plot showing distribution of fold enrichment values for the DIP-seq peaks with fSBM, mSBM, and others. (I) IGV images showing ChIP-seq peaks on the upstream of PbSIP2 target genes. Red and pink bars indicate location of the fSBM and mSBM, respectively. Scales are indicated in square brackets. (J) Histogram showing the number of genes against the log2(fold change) value in the differential expression analysis between pbsip2-cKORapa− and pbsip2-cKORapa+. The percentages of genes associated with SBM are indicated on the graph.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

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

The self-check needs the paper’s full text.

Citations

Cites

Metadata

Authors
Nishi, T., Kaneko, I., & Yuda, M.
Affiliations
Mie University (JP)
Corresponding author
Yuda, Masao
ORCID
Venue
Science Advances
Publisher
American Association for the Advancement of Science (AAAS)
Published
2025-03-21
Volume
11
Issue
12
Pages
eads5458-eads5458
ISSN
2375-2548 (ISSN-L)
DOI
10.1126/sciadv.ads5458
PMC
PMC11927625
OpenAlex
W4408718230
Cited by
5
References
55
Citations by year
2026: 3 · 2025: 2
FWCI
4.53
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • Science Advances (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
Japan Society for the Promotion of Science (24K10187, 23K27400)
Topics
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Malaria Research and Control
  • Life Sciences › Immunology and Microbiology › Parasitology › Toxoplasma gondii Research Studies
  • Life Sciences › Immunology and Microbiology › Parasitology › Bird parasitology and diseases
Keywords
SiP2, Plasmodium berghei, schizogony, AP2 transcription factor, erythrocyte invasion, transcriptional regulation, cis-regulatory elements, conditional gene disruption, malaria pathogenesis
MeSH terms
Animals, Erythrocytes, Gene Expression Regulation, Malaria, Plasmodium berghei, Transcription Factors, Protozoan Proteins, Mice, Merozoites

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.

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

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.