Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins
Open accessThe DOI matches its Crossref record; 36 of the 36 references it cites were found in a registry or an index; none of the 71 sources its claims cite is checked for support yet.
Abstract
Determining the origin of recurrent Plasmodium vivax infections is critical for quantifying the public health burden of the hypnozoite reservoir, understanding transmission potential, and evaluating anti-relapse antimalarial efficacy. In Ethiopia, P. vivax is a major contributor to malaria burden, but the relative contribution of relapse to recurrent infections remains poorly quantified. We performed microhaplotype genotyping using a panel of 93 genome-wide markers and applied a probabilistic recurrence classification framework to recurrent P. vivax samples collected in two Ethiopian radical cure trials (IMPROV and OPRA). Genotyping was conducted on parasite isolates collected at enrolment and subsequent recurrences, with participants monitored over a 12-month follow-up period (IMPROV: 604 samples from 444 participants; OPRA: 621 samples from 342 participants). Recurrent infections were classified using a Bayesian hierarchical model implemented in Pv3Rs , which estimated marginal posterior probabilities of relapse, reinfection, or recrudescence based on genotyping data and background population dynamics. Priors were structured across biologically motivated time windows reflecting the natural history of P. vivax infection in Ethiopia and anticipated drug efficacy. Recurrences with a probability of relapse (p(L)) ≥ 0.75 were classified as relapses. In the absence of primaquine, probable relapse was the dominant classification for primary recurrent infections in both trials (75.5%; 102/135) and for consecutive recurrent infections within individuals (75.2%; 204/271). Primaquine substantially reduced the proportion of participants experiencing recurrent infections attributable to probable relapse, from 57.3% to 1.7% in IMPROV and from 60.2% to 9.5% in OPRA. Polyclonal infection at enrolment was independently associated with increased odds of recurrence and a higher rate of probable relapse in OPRA, with a consistent but non-significant trend observed in IMPROV. These findings provide direct molecular evidence that the hypnozoite reservoir is the dominant driver of recurrent P. vivax infections in Ethiopia in the absence of radical cure. The marked reduction in probable relapse following primaquine treatment highlights the critical importance of expanding access to safe and effective radical cure strategies in Ethiopia and other comparable high-transmission settings.
Overview
CoChat Summary
TL;DR
Microhaplotype sequencing and Bayesian recurrence classification in two Ethiopian *P. vivax* trials show that, without primaquine, probable relapse dominates recurrent infections, while primaquine sharply reduces relapse-attributable recurrence. Polyclonal infection at enrolment was associated with higher recurrence and relapse risk in OPRA, with a non-significant trend in IMPROV.
Key findings
- In the absence of primaquine, probable relapse was the dominant classification for primary recurrent infections in both trials (75.5%; 102/135) and for consecutive recurrent infections within individuals (75.2%; 204/271).
- Primaquine substantially reduced the proportion of participants experiencing recurrent infections attributable to probable relapse, from 57.3% to 1.7% in IMPROV and from 60.2% to 9.5% in OPRA.
- In OPRA non-primaquine arms, polyclonal infection at enrolment was associated with recurrence (74.0% vs 55.0%; OR = 2.35, 95% CI 1.14–4.83, p = 0.020) and a higher count of probable relapse episodes (ratio = 1.39, 95% CI 1.01–1.92, p = 0.044); IMPROV showed a consistent but non-significant trend (OR = 1.63, 95% CI 0.70–3.81, p = 0.26; relapse ratio = 1.14, 95% CI 0.61–2.14, p = 0.68).
- Within-host identity-by-descent was lower in primaquine arms; for example, IMPROV control mean consecutive-pair IBD was 58% vs 23% in PQ arms (p = 1.1×10⁻⁵), and OPRA mean IBD was 75% in AL, 69% in CQ, and 55% in PQ arms (Kruskal-Wallis H = 9.5, p = 0.0086).
Methods
- Microhaplotype genotyping used the vivaxGEN *P. vivax* 93-amplicon panel on parasite isolates collected at enrolment and subsequent recurrences over 12-month follow-up in two Ethiopian radical cure trials: IMPROV (604 samples from 444 participants) and OPRA (621 samples from 342 participants).
- Within-host diversity was estimated with MOIRE effective multiplicity of infection (eMOI), pairwise relatedness with Dcifer identity-by-descent, and recurrence origin with the Pv3Rs Bayesian hierarchical model, which returned marginal posterior probabilities of relapse, reinfection, and recrudescence using time-window priors; relapse was classified as p(L) ≥ 0.75, reinfection as p(L) ≤ 0.25, and values between 0.25 and 0.75 as indeterminate.
- Association between Day 0 clonality and recurrence or probable relapse was assessed with logistic regression restricted to non-primaquine control arms and fitted separately per trial.
Limitations
- Stated: sampling bias because not every recurrence was genotyped; OPRA primaquine adherence was semi-supervised for the initial episode and unsupervised for subsequent recurrences, likely attenuating efficacy; the conservative reinfection prior (λ = 0.002/day) makes probable relapse estimates lower-bound.
- Stated: the impact of polyclonality on classification could not be assessed robustly in IMPROV due to limited monoclonal pairs and more Pv3Rs computational failures; Pv3Rs’ two-parent assumption can misclassify half-siblings; fixed time-window priors were used, and a dedicated time-to-event model would allow more flexibility.
- Evident from the text: transmission potential could not be directly assessed, and fully resolving genetic identity within polyclonal recurrences would require haplotype-deconvolved whole-genome data, which was beyond the study’s scope.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures









![Supplementary Figure S4: Identity-by-descent (IBD) network plot of baseline parasite populations within IMPROV and OPRA (IBD Thresholds [left to right]: 0.125, 0.25, 0.50, 0.95). (A) All Day 0 samples from IMPROV study sites: Arba Minch and Metehara. (B) All Day 0 samples from OPRA study sites: Batu and Bishoftu. (C) All Day 0 samples from both IMPROV and OPRA. (D) All monoclonal (eMOI ≤ 1.1) Day 0 samples from both IMPROV and OPRA.](https://figures.cochat.ai/paper-figures/0e10111aa61b4c8b93278b1ee5e0ebf2/i-plasmodium-vivax-i-relapse-reservoirs-in-ethiopia-microhaplotype-sequencing-figure-10.png)








Verification
Claims the paper makes
0 Valid · 0 Not valid · 41 Not checked · 144 Own results and interpretations
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“There were significantly fewer classified probable relapses in the PQ arms at 12.9% (4/31), compared to 80.6% (25/31) classified as reinfection and 6.5% (2/31) as indeterminate.”
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Relapses driving primary recurrent infections in the absence of primaquine
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Primaquine substantially reduced the proportion of participants experiencing recurrent infections attributable to probable relapse, from 57.3% to 1.7% in IMPROV”
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Abstract
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Relapse estimates in the PQ arms of OPRA (62.5%; 25/40) were higher than that in the PQ arms of IMPROV (15.2%; 5/33).”
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Relapse is the dominant mechanism of recurrence in the absence of radical cure
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The classification proportions differed significantly across the three OPRA arms (χ 2 = 12.9, df = 4, p = 0.0117).”
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Supplementary
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and for consecutive recurrent infections within individuals (75.2%; 204/271)”
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Abstract
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
101 more claims
Create a free account to see all 185 claims and to verify the 71 references still open.
- It may state two separate claims as one. 57
- Probably not a substantial claim. 29
- The quote only partly states this claim. 4
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.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Abstract
“Primaquine substantially reduced the proportion of participants experiencing recurrent infections attributable to probable relapse, from 57.3% to 1.7% in IMPROV and from 60.2% to 9.5% in OPRA.”
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Relapses driving primary recurrent infections in the absence of primaquine
“In IMPROV, 57.3% (43/75) of participants in the control arm had a clinical recurrence of P. vivax during the 12-month follow-up period”
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Relapse is the dominant mechanism of recurrence in the absence of radical cure
“Of the 147 participants treated with PQ, the majority of recurrences were attributed to probable reinfection with only 9.5% (14/147) classified as probable relapse.”
§ Long title (250 characters max): Plasmodium vivax relapse reservoirs in Ethiopia: microhaplotype sequencing informs recurrent infection origins > Relapse is the dominant mechanism of recurrence in the absence of radical cure
“In OPRA, 75.6% (90/119) of AL pairs and 75.0%(78/104) of CQ pairs were classified as probable relapses, compared to 62.5% (25/40) of PQ arm pairs”
Citations
Cites
DADA2: High–resolution sample inference from Illumina amplicon data
Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., & Holmes, S. P. (2016) · Nature Methods · doi:10.1038/nmeth.3869 · Cited by 35,698
Glucose-6-phosphate dehydrogenase deficiency
Cappellini, M., & Fiorelli, G. (2008) · The Lancet · doi:10.1016/s0140-6736(08)60073-2 · Cited by 1,590
Vivax Malaria: Neglected and Not Benign
Price, R. N., Tjitra, E., Guerra, C. A., Yeung, S., White, N. J., & Anstey, N. M. (2007) · The American Journal of Tropical Medicine and Hygiene · doi:10.4269/ajtmh.2007.77.79 · Cited by 838
Self-citationDeterminants of relapse periodicity in Plasmodium vivax malaria
White, N. J. (2011) · Malaria Journal · doi:10.1186/1475-2875-10-297 · Cited by 679
The IMPROV Study Group (2015) · BMC Infectious Diseases · doi:10.1186/s12879-015-1276-2 · Cited by 277
fastp 1.0: An ultra‐fast all‐round tool for FASTQ data quality control and preprocessing
Chen, S. (2025) · iMeta · doi:10.1002/imt2.70078 · Cited by 258
Taylor, W. R. J., Thriemer, K., von Seidlein, L., Yuentrakul, P., Assawariyathipat, T., Assefa, A., Auburn, S., Chand, K., Chau, N. H., Cheah, P. Y., et al. (2019) · The Lancet · doi:10.1016/s0140-6736(19)31285-1 · Cited by 178
Self-citationWhite, N. J., & Imwong, M. (2012) · Advances in Parasitology · doi:10.1016/b978-0-12-397900-1.00002-5 · Cited by 163
Commons, R. J., Simpson, J. A., Watson, J., White, N. J., & Price, R. N. (2020) · The American Journal of Tropical Medicine and Hygiene · doi:10.4269/ajtmh.20-0186 · Cited by 147
Self-citationCharacterization of Within-Host Plasmodium falciparum Diversity Using Next-Generation Sequence Data
Auburn, S., Campino, S., Miotto, O., Djimde, A. A., Zongo, I., Manske, M., Maslen, G., Mangano, V., Alcock, D., MacInnis, B., et al. (2012) · PLoS ONE · doi:10.1371/journal.pone.0032891 · Cited by 142
Self-citationResolving the cause of recurrent Plasmodium vivax malaria probabilistically
Taylor, A. R., Watson, J. A., Chu, C. S., Puaprasert, K., Duanguppama, J., Day, N. P. J., Nosten, F., Neafsey, D. E., Buckee, C. O., Imwong, M., et al. (2019) · Nature Communications · doi:10.1038/s41467-019-13412-x · Cited by 118
Self-citationDiscovery of the hypnozoite and a new theory of malarial relapse
Krotoski, W. A. (1985) · Transactions of the Royal Society of Tropical Medicine and Hygiene · doi:10.1016/0035-9203(85)90221-4 · Cited by 116
The prevention and treatment of Plasmodium vivax malaria
Chu, C. S., & White, N. J. (2021) · PLOS Medicine · doi:10.1371/journal.pmed.1003561 · Cited by 88
Why Is Plasmodium vivax a Neglected Tropical Disease?
Carlton, J. M., Sina, B. J., & Adams, J. H. (2011) · PLoS Neglected Tropical Diseases · doi:10.1371/journal.pntd.0001160 · Cited by 86
Bright, A. T., Manary, M. J., Tewhey, R., Arango, E. M., Wang, T., Schork, N. J., Yanow, S. K., & Winzeler, E. A. (2014) · PLoS Neglected Tropical Diseases · doi:10.1371/journal.pntd.0002882 · Cited by 84
Abreha, T., Hwang, J., Thriemer, K., Tadesse, Y., Girma, S., Melaku, Z., Assef, A., Kassa, M., Chatfield, M. D., Landman, K. Z., et al. (2017) · PLOS Medicine · doi:10.1371/journal.pmed.1002299 · Cited by 82
Self-citationTessema, S. K., Hathaway, N. J., Teyssier, N. B., Murphy, M., Chen, A., Aydemir, O., Duarte, E. M., Simone, W., Colborn, J., Saute, F., et al. (2022) · The Journal of Infectious Diseases · doi:10.1093/infdis/jiaa527 · Cited by 74
Dcifer: an IBD-based method to calculate genetic distance between polyclonal infections
Gerlovina, I., Gerlovin, B., Rodríguez-Barraquer, I., & Greenhouse, B. (2022) · Genetics · doi:10.1093/genetics/iyac126 · Cited by 71
Modelling primaquine-induced haemolysis in G6PD deficiency
Watson, J., Taylor, W. R., Menard, D., Kheng, S., & White, N. J. (2017) · eLife · doi:10.7554/elife.23061 · Cited by 59
Prevention and treatment of vivax malaria
Baird, J. K., Schwartz, E., & Hoffman, S. L. (2007) · Current Infectious Disease Reports · doi:10.1007/s11908-007-0021-4 · Cited by 36
Devine, A., Battle, K. E., Meagher, N., Howes, R. E., Dini, S., Gething, P. W., Simpson, J. A., Price, R. N., & Lubell, Y. (2021) · PLOS Medicine · doi:10.1371/journal.pmed.1003614 · Cited by 33
Self-citationMurphy, M., & Greenhouse, B. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btae619 · Cited by 33
Ross, A., Koepfli, C., Schoepflin, S., Timinao, L., Siba, P., Smith, T., Mueller, I., Felger, I., & Tanner, M. (2016) · PLOS Neglected Tropical Diseases · doi:10.1371/journal.pntd.0004582 · Cited by 32
Siegel, S. V., Trimarsanto, H., Amato, R., Murie, K., Taylor, A. R., Sutanto, E., Kleinecke, M., Whitton, G., Watson, J. A., Imwong, M., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-51015-3 · Cited by 30
Self-citationPopulation heterogeneity in Plasmodium vivax relapse risk
Stadler, E., Cromer, D., Mehra, S., Adekunle, A. I., Flegg, J. A., Anstey, N. M., Watson, J. A., Chu, C. S., Mueller, I., Robinson, L. J., et al. (2022) · PLOS Neglected Tropical Diseases · doi:10.1371/journal.pntd.0010990 · Cited by 20
Baseline malaria prevalence at the targeted pre-elimination districts in Ethiopia
Nega, D., Abera, A., Gidey, B., Mekasha, S., Abebe, A., Dillu, D., Mehari, D., Assefa, G., Hailu, S., Haile, M., et al. (2021) · BMC Public Health · doi:10.1186/s12889-021-12036-5 · Cited by 19
Self-citationMicrohaplotype deep sequencing assays to capture Plasmodium vivax infection lineages
Kleinecke, M., Sutanto, E., Rumaseb, A., Hoon, K. S., Trimarsanto, H., Osborne, A., Manrique, P., Peters, T., Hawkes, D., Benavente, E. D., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-62357-x · Cited by 19
Self-citationHigh-throughput genotyping of Plasmodium vivax in the Peruvian Amazon via molecular inversion probes
Popkin-Hall, Z. R., Niaré, K., Crudale, R., Simkin, A., Fola, A. A., Sanchez, J. F., Pannebaker, D. L., Giesbrecht, D. J., Kim, I. E., Aydemir, Ö., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-54731-y · Cited by 11
Shumbej, T., Jemal, A., Worku, A., Bekele, F., & Weldesenbet, H. (2019) · BMC Infectious Diseases · doi:10.1186/s12879-019-4055-7 · Cited by 11
Thriemer, K., Degaga, T. S., Christian, M., Alam, M. S., Ley, B., Hossain, M. S., Kibria, M. G., Tego, T. T., Abate, D. T., Weston, S., et al. (2022) · Trials · doi:10.1186/s13063-022-06364-z · Cited by 8
Self-citationPv3Rs: Plasmodium vivax relapse, recrudescence, and reinfection statistical genetic inference
Foo, Y. S., White, M. T., & Taylor, A. R. (2026) · Bioinformatics · doi:10.1093/bioinformatics/btaf643 · Cited by 6
Self-citationRosado, J., Han, J., Obadia, T., Munro, J., Traore, Z., Schoffer, K., Brewster, J., Bourke, C., Vinetz, J. M., Taylor, A. R., et al. (2026) · iScience · doi:10.1016/j.isci.2026.115799 · Cited by 3
Self-citationPopovici, J., Seng, D., Ko, K., Orban, A., Heng, S., Tat, B., Feufack-Donfack, L., Eng, V., Tebben, K., Khim, N., et al. (2026) · doi:10.21203/rs.3.rs-9199901/v1 · Cited by 2
Degaga, T. S., Sime, H., Ruybal-Pesántez, S., Osborne, A., Mwaura, M., Bamboro, S. A., Mnjala, H., Mekonnen, G., Ermias, S., Fantahun, A., et al. (2026) · Malaria Journal · doi:10.1186/s12936-026-06050-4 · Cited by 1
Self-citationRevisiting Plasmodium vivax molecular correction
Taylor, A. R., Foo, Y. S., & White, M. T. (2026) · doi:10.64898/2026.06.02.26354709 · Cited by 1
Self-citationP. vivax microhaplotype deep sequencing assays v2Preprint
Rumaseb, A., Peter, T., Osborne, A., Trimarsanto, H., Kleinecke, M., Hawkes, D. B., & Auburn, S. (2026) · doi:10.17504/protocols.io.x54v92boml3e/v2 · Cited by 1
Self-citation
Research works printed but not linked: 1
- Bakker JD. Mantel Test. 2026 [cited 27 Aug 2026]. Available: https://uw.pressbooks.pub/appliedmultivariatestatistics/chapter/mantel-test/
Cited by
Still being checked.
Metadata
- Authors
- Osborne, A. A., Ruybal-Pesantez, S., Rai, A., Trimarsanto, H., Hoon, K. S., Piera, K., Kleinecke, M., Rumaseb, A., Sutanto, E., Thriemer, K., Degaga, T. S., Hwang, J., Assefa, A., Woyessa, A., Hailu, A., Taylor, A., Day, N. P., Price, R. N., & Auburn, S.
- Affiliations
- Charles Darwin University (AU)
- Barcelona Institute for Global Health, ISGlobal, Campus Clínic Hospital Clínic, Universitat de Barcelona, Barcelona, Spain
- Centers for Disease Control and Prevention (US)
- University of North Carolina at Chapel Hill (US)
- Ethiopian Public Health Institute (ET)
- Institut Pasteur (FR)
- Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, UK
- Mahidol University (TH)
- Corresponding author
- Osborne, Ashley A.
- ORCID
- Venue
- bioRxiv (Cold Spring Harbor Laboratory)
- Publisher
- openRxiv
- Published
- 2026-11-13
- OpenAlex
- W7225986520
- Semantic Scholar
- cb2fd5ad758adff36d1ec6bd0d0d2152f2c20fe6
- Cited by
- 0
- References
- 36
- Type
- posted content
- Language
- en
- Open access
- green
- doi.org
- Licence
- Locations
- bioRxiv (Cold Spring Harbor Laboratory) (submittedVersion, cc-by)
- Funders
- National Health and Medical Research Council of Australia (Investigator Grant 2025377)
- Topics
- Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Malaria Research and Control
- Life Sciences › Immunology and Microbiology › Parasitology › Bird parasitology and diseases
- Life Sciences › Immunology and Microbiology › Parasitology › Toxoplasma gondii Research Studies
- Keywords
- Plasmodium vivax, relapse, primaquine, radical cure, polyclonal infection, reinfection, recrudescence, Ethiopia, Bayesian hierarchical model
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.
A language model drew the claims from the paper on 2026-11-21, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-11-21 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
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.