Correlates of protection against African swine fever virus identified by a systems immunology approach
- Institute of Virology and Immunology IVI, Switzerland
- Department of Infectious Diseases and Pathobiology, Vetsuisse Faculty, University of Bern, Switzerland
- Graduate School for Cellular and Biomedical Sciences, University of Bern, Switzerland
- Multidisciplinary Center for Infectious Diseases, University of Bern, Switzerland
- Department of Virology, Wageningen Bioveterinary Research, Wageningen University and Research, Netherlands
Figures
Figure 1 with 3 supplements
Outcomes of African swine fever virus (ASFV) immunization and challenge in specific pathogen-free (SPF) and farm pigs.
(A) Overview of experiment layout marking time points of sample collection and performed immunological readouts (dark rectangles indicate collected data). Figure was created with BioRender.com. (B, F) Rectal temperatures and clinical scores were recorded daily after immunization with the attenuated Estonia 2014 strain and after challenge with the highly virulent Armenia 2008 strain. (C, H) Viral DNA levels in blood and serum were measured by qPCR. (D, I) Organs were collected at 17 days post-infection (dpi) (D) and at 7, 10, 13, and 26 days post-challenge (dpc) (I) for quantification of viral DNA levels (SM LN – submandibular lymph node; GH LN – gastrohepatic lymph node). (E) Seroconversion was analyzed by competitive ELISA. For B, C, E, and F, the data points represent values from individual animals, and the lines show means of the groups. For D, H, and I, the data points represent values for individual animals. There were n=9 pigs per group in B, C, and E, and n=3 in D. In F and G, there were n=5 pigs per group at 0 dpc. In H, at 4 and 7 dpc, n=5 pigs in farm and SPF groups; at 11 dpc, n=2 in farm group and n=5 in SPF group; and at 18 and 25 dpc, n=2 in farm group and n=4 in SPF group. In I, n=5 pigs in farm and SPF groups. For B and F, the differences in temperature and clinical scores between two groups were analyzed by comparing areas under the curves (AUCs) with unpaired t-test. For C, D, and E, the differences between farm and SPF groups were analyzed at each time point by unpaired t-test with Holm-Šídák’s correction for multiple comparisons; *p<0.05. For G, the survival rates in two groups were compared by Log-rank test. For H and I, the differences in viral loads between protected and non-protected animals were analyzed by unpaired t-test with Holm-Šídák’s correction for multiple comparisons; §p<0.05; §§§p<0.001.
Figure 1—figure supplement 1
Hematologic profiles after immunization with the Estonia 2014 strain.
Leukocyte, erythrocyte, and platelet counts with data points representing individual animals and lines showing the means of farm and specific pathogen-free (SPF) groups. There were n=9 pigs in each group. Differences to baseline measurements (day –5) for farm and SPF groups were analyzed by mixed-effects analysis with Dunnett’s multiple comparisons test (blue hashtags – farm group, pink hashtags – SPF group); #p<0.05; ##p<0.01; ###p<0.001; ####p<0.0001. Differences between farm and SPF groups were analyzed at each time point by unpaired t-test with Holm-Šidák’s correction for multiple comparisons; *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.
Figure 1—figure supplement 2
Viral DNA levels in serum after challenge.
Quantification was performed by qPCR. Data points represent values for individual animals remaining in the experiment at various time points post-challenge. At 4 and 7 days post-challenge (dpc), n=5 pigs in farm and specific pathogen-free (SPF) groups; at 11 dpc, n=2 in farm group and n=5 in SPF group; and at 18 and 25 dpc, n=2 in farm group and n=4 in SPF group.
Figure 1—figure supplement 3
Hematologic profiles after challenge with the Armenia 2008 strain.
Leukocyte, erythrocyte, and platelet counts of farm and specific pathogen-free (SPF) pigs. Data points represent values for individual animals; lines show the means of the groups. At 4 and 7 days post-challenge (dpc), n=5 pigs in farm and SPF groups; at 11 dpc, n=2 in farm group and n=5 in SPF group; and at 18 and 26 dpc, n=2 in farm group and n=4 in SPF group. Differences between protected and non-protected animals were analyzed by unpaired t-test with Holm-Šídák’s correction for multiple comparisons; §p<0.05.
Figure 2 with 3 supplements
Cytokine responses following immunization and challenge and their correlation with protection.
(A, C) Serum cytokine levels were measured by ELISA. (B, D) Dot plots show correlations between cytokine levels and clinical scores after challenge (r – correlation coefficient; dark red indicates a negative correlation with protection, which reflects poor clinical outcomes, while gray indicates a positive correlation with protection; correlations with p<0.1 are outlined with black circles). For A, the data points represent values for individual animals and the lines show means of the groups. For C, the data points represent values for individual animals. In A, there were n=8 farm pigs and n=9 specific pathogen-free (SPF) pigs, and in C, at –9, 4, and 7 days post-challenge (dpc), n=5 pigs in farm and SPF groups, and at 11 dpc, n=2 in farm group and n=5 in SPF group. In A, the differences to baseline measurements (day –5) for farm and SPF pigs were analyzed by two-way ANOVA with Dunnett’s multiple comparisons test (blue hashtags – farm group, pink hashtags – SPF group); #p<0.05; ##p<0.01; ###p<0.001; ####p<0.0001. Differences between farm and SPF groups were analyzed at each time point by unpaired t-test with Holm-Šidák’s correction for multiple comparisons; *p<0.05; **p<0.01; ****p<0.0001. In C, the differences to baseline measurements (day –9) for farm, SPF, protected and non-protected groups were analyzed by mixed-effects analysis with Dunnett’s multiple comparisons test (pink hashtags – SPF group, Ø – non-protected animals); #p<0.05; Øp<0.05; ØØp<0.01; ØØØØp<0.0001. Differences between farm and SPF groups, as well as protected and non-protected animals, were analyzed at each time point by unpaired t-test with Holm-Šidák’s correction for multiple comparisons.
Figure 2—figure supplement 1
Clinical score values for area under the curve (AUC) calculations.
(A) The clinical scores post-challenge, used to calculate the AUC for each animal, are shown. The lines were prolonged until 16 days post-challenge (dpc) for the animals reaching clinical end points to obtain an expected minimum AUC. (B) AUC values for farm and specific pathogen-free (SPF) pigs used for correlations.
Figure 2—figure supplement 2
Correlation of serum cytokine levels post-immunization with clinical outcomes (areas under the curves [AUCs]).
For each group, n=5 pigs were included in the analysis.
Figure 2—figure supplement 3
Correlation of serum cytokine levels post-challenge with clinical outcomes (areas under the curves [AUCs]).
At –9, 4, and 7 days post-challenge (dpc), n=5 pigs in farm and specific pathogen-free (SPF) groups, and at 11 dpc, n=2 in farm group and n=5 in SPF group.
Figure 3 with 5 supplements
Cellular immune responses before and after challenge and their correlation with protection.
(A) IFN-γ release from freshly collected peripheral blood mononuclear cells (PBMCs) upon African swine fever virus (ASFV) restimulation was measured by ELISpot before and after challenge (SFC – spot-forming cell). Correlations between IFN-γ response and clinical scores after challenge are shown on the plots. (B) T-cell responses upon ASFV in vitro restimulation were analyzed by intracellular cytokine staining. (C) Correlations between the frequencies of cytokine-producing T-cell subsets and clinical scores after challenge are displayed in dot plots (r – correlation coefficient; dark red indicates a negative correlation with protection, which reflects poor clinical outcomes, while gray indicates a positive correlation with protection; correlations with p<0.05 are outlined with black circles). In A, the data points represent values for individual animals. Measurements were made in triplicate; bars represent mean values for each animal. In B, the data points represent values for individual animals. At –9 and 7 days post-challenge (dpc), n=5 pigs in farm and specific pathogen-free (SPF) groups, and at 18 and 26 dpc, n=2 in farm group and n=4 in SPF group. Differences between farm and SPF groups (*), as well as between protected and non-protected animals (§), were analyzed at each time point by unpaired t-test with Holm-Šidák’s correction for multiple comparisons; *p<0.05; **p<0.01; §p<0.05; §§p<0.01.
Figure 3—figure supplement 1
Gating strategy for analysis of intracellular cytokine responses following in vitro restimulation of freshly collected peripheral blood mononuclear cells (PBMCs) with live African swine fever virus (ASFV).
Figure 3—figure supplement 2
Cellular immune responses from effector γδ T cells and NK cells during the challenge phase.
Freshly collected peripheral blood mononuclear cells (PBMCs) were restimulated with African swine fever virus (ASFV). Supernatant from mock-infected macrophages served as negative control; obtained percentages were subtracted. Data points represent values for individual animals. At –9 and 7 days post-challenge (dpc), n=5 pigs in farm and specific pathogen-free (SPF) groups, and at 18 and 26 dpc, n=2 in farm group and n=4 in SPF group. Differences between protected and non-protected animals were analyzed by unpaired t-test with Holm-Šídák’s correction for multiple comparisons; §p<0.05.
Figure 3—figure supplement 3
Correlation of cytokine responses in three cell subsets post-challenge with clinical outcomes (areas under the curves [AUCs]).
At –9 and 7 days post-challenge (dpc), n=5 pigs in farm and specific pathogen-free (SPF) groups, and at 18 and 26 dpc, n=2 in farm group and n=4 in SPF group.
Figure 3—figure supplement 4
Correlation of cytokine responses in two cell subsets post-challenge with clinical outcomes (areas under the curves [AUCs]).
At –9 and 7 days post-challenge (dpc), n=5 pigs in farm and specific pathogen-free (SPF) groups, and at 18 and 26 dpc, n=2 in farm group and n=4 in SPF group.
Figure 3—figure supplement 5
Swine leukocyte antigen (SLA) haplotyping.
(A) Distribution of alleles for MHC (SLA) class I and class II in farm and specific pathogen-free (SPF) pigs. (B) Allele diversity and composition of haplotypes in farm and SPF pigs.
Figure 4 with 2 supplements
Antigen-specific antibody responses.
(A) Semiquantitative anti-p72 antibody levels expressed as percent inhibition in samples collected before and after challenge. (B) Comparison of antibody titers defined by hemadsorption inhibition assay (HADIA). Differences between farm and specific pathogen-free (SPF) groups were analyzed at each time point by unpaired t-test with Holm-Šidák’s correction for multiple comparisons; *p<0.05; **p<0.01. Both farm and SPF groups had n=5 pigs each.
Figure 4—figure supplement 1
Correlation analysis between anti-p72 antibody titers and protection.
(A) Two serum samples from a randomly selected protected specific pathogen-free (SPF) and a non-protected farm pig were serially diluted to define, by competitive ELISA, the non-saturating concentration estimated to provide 20–80% inhibition. The dilution factor of 1:1250 was chosen to test all serum samples. Measurements were made in duplicate; bars represent mean values with standard deviation (SD). (B) Correlation of antibody levels expressed as % inhibition with clinical outcomes.
Figure 4—figure supplement 2
Correlation analysis between hemadsorption-inhibiting antibody titers and protection.
(A) Workflow of hemadsorption inhibition assay (HADIA) for determining antibody titers. Figure was created with BioRender.com. (B) Correlation of antibody titers with clinical outcomes.
Figure 5 with 2 supplements
Blood leukocyte transcriptomic profiles after immunization with the Estonia 2014 strain.
For each group of animals, comparisons were made against the baseline measurements (day 0). Pre-ranked list of differentially expressed genes (DEGs) was subjected to Gene Set Enrichment Analysis (GSEA) using blood transcription modules (BTMs) as gene sets. (A, B) Dot plots show innate and adaptive BTMs, respectively. Dot size depicts the q-value (FDR), while color represents the normalized enrichment score (NES). FDR<0.05 was selected as a cutoff.
Figure 5—figure supplement 1
Blood leukocyte transcriptomic profiles after immunization with the Estonia 2014 strain.
For each group of animals, comparisons were made against the baseline measurements (day 0). Pre-ranked list of differentially expressed genes (DEGs) was subjected to Gene Set Enrichment Analysis (GSEA) using blood transcription modules (BTMs) as gene sets. (A, B) Dot plots show non-classified and to-be-annotated (TBA) BTMs, respectively. Dot size depicts the q-value (FDR), while color represents the normalized enrichment score (NES). FDR<0.05 was selected as a cutoff.
Figure 5—figure supplement 2
Lymphoid tissue transcriptomic profiles after immunization with the Estonia 2014 strain.
Gastrohepatic lymph nodes (LNs) and spleens were collected at 17 days post-infection (dpi). For each group of animals, comparisons were made against the mock group. In addition, farm and specific pathogen-free (SPF) pigs were compared between each other. Pre-ranked list of differentially expressed genes (DEGs) was subjected to Gene Set Enrichment Analysis (GSEA) using blood transcription modules (BTMs) as gene sets. (A, B) Dot plots show innate and adaptive BTMs, respectively. Dot size depicts the q-value (FDR), while color represents the normalized enrichment score (NES). FDR<0.05 was selected as a cutoff.
Figure 6 with 4 supplements
Correlation of relative blood transcription module (BTM) expression after Estonia 2014 immunization with protection.
Read counts after bulk RNA sequencing were normalized by the baseline values (day 0). Gene Set Variation Analysis (GSVA) was then performed to calculate normalized enrichment score (NES) values for individual animals using BTMs as gene sets. Obtained NES values were correlated with area under the curve (AUC) values of the challenge clinical outcomes. In A and B, the dot plots show innate and adaptive BTMs, respectively. Dot size depicts the p-value, and color represents the correlation coefficient r (dark red indicates a negative correlation with protection, which reflects poor clinical outcomes, while gray indicates a positive correlation with protection). Correlations with p<0.05 are outlined with black circles.
Figure 6—figure supplement 1
Heatmap showing relative blood transcription module (BTM) expression in blood leukocytes at 4 days post-infection (dpi).
Gene expression counts were obtained after bulk RNA sequencing and normalized using day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 6—figure supplement 2
Heatmap showing relative blood transcription module (BTM) expression in blood leukocytes at 7 days post-infection (dpi).
Gene expression counts were obtained after bulk RNA-seq and normalized using day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 6—figure supplement 3
Heatmap showing relative blood transcription module (BTM) expression in blood leukocytes at 11 days post-infection (dpi).
Gene expression counts were obtained after bulk RNA sequencing and normalized using day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 6—figure supplement 4
Heatmap showing relative blood transcription module (BTM) expression in blood leukocytes at 15 days post-infection (dpi).
Gene expression counts were obtained after bulk RNA sequencing and normalized using day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 7 with 5 supplements
Correlation of absolute blood transcription module (BTM) expression after Estonia 2014 immunization with protection.
Gene Set Variation Analysis (GSVA) was performed on non-normalized read counts to calculate normalized enrichment score (NES) values for individual animals using BTMs as gene sets. Obtained NES values were correlated with area under the curve (AUC) values of the challenge clinical outcomes. In A and B, the dot plots show innate and adaptive BTMs, respectively. Dot size depicts the p-value, and color represents the correlation coefficient r (dark red indicates a negative correlation with protection, which reflects poor clinical outcomes, while gray indicates a positive correlation with protection). Correlations with p<0.05 are outlined with black circles.
Figure 7—figure supplement 1
Heatmap showing absolute blood transcription module (BTM) expression in blood leukocytes at baseline (day 0).
Gene expression counts were obtained after bulk RNA sequencing and used without normalization. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmap. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 7—figure supplement 2
Heatmap showing absolute blood transcription module (BTM) expression in blood leukocytes at 4 days post-infection (dpi).
Gene expression counts were obtained after bulk RNA sequencing and used without normalization to day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 7—figure supplement 3
Heatmap showing absolute blood transcription module (BTM) expression in blood leukocytes at 7 days post-infection (dpi).
Gene expression counts were obtained after bulk RNA sequencing and used without normalization to day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 7—figure supplement 4
Heatmap showing absolute blood transcription module (BTM) expression in blood leukocytes at 11 days post-infection (dpi).
Gene expression counts were obtained after bulk RNA sequencing and used without normalization to day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 7—figure supplement 5
Heatmap showing absolute blood transcription module (BTM) expression in blood leukocytes at 15 days post-infection (dpi).
Gene expression counts were obtained after bulk RNA sequencing and used without normalization to day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side. Pigs 12–20: farm group; pigs 1963–1975: specific pathogen-free (SPF) group.
Figure 8
Kinetics of average normalized enrichment score (NES) of blood transcription module (BTM) families after Estonia 2014 immunization.
NES values were calculated by Gene Set Variation Analysis (GSVA) for five farm and five specific pathogen-free (SPF) pigs using non-normalized read counts. At each time point, NES values of BTMs belonging to one family were used to determine average family scores. Plots demonstrate average NES values for each family related to either innate (left column) or adaptive immunity (right column). Both farm and SPF groups had n=5 pigs each. Differences between protected and non-protected farm pigs were analyzed by unpaired t-test with Holm-Šídák’s correction for multiple comparisons; §p<0.05.
Figure 9 with 4 supplements
Correlation of absolute blood transcription module (BTM) expression after Armenia 2008 challenge with protection.
Gene Set Variation Analysis (GSVA) was performed on non-normalized read counts to calculate normalized enrichment score (NES) values for individual animals using BTMs as gene sets. Obtained NES values were correlated with area under the curve (AUC) values of the challenge clinical outcomes. In A and B, the dot plots show innate and adaptive BTMs, respectively. Dot size depicts the p-value, and color represents the correlation coefficient r (dark red indicates a negative correlation with protection, which reflects poor clinical outcomes, while gray indicates a positive correlation with protection). Correlations with p<0.05 are outlined with black circles.
Figure 9—figure supplement 1
Heatmap showing absolute blood transcription module (BTM) expression in blood leukocytes at 4 days post-challenge (dpc).
Gene expression counts were obtained after bulk RNA sequencing and used without normalization to day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side.
Figure 9—figure supplement 2
Heatmap showing absolute blood transcription module (BTM) expression in blood leukocytes at 7 days post-challenge (dpc).
Gene expression counts were obtained after bulk RNA sequencing and used without normalization to day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side.
Figure 9—figure supplement 3
Heatmap showing absolute blood transcription module (BTM) expression in blood leukocytes at 11 days post-challenge (dpc).
Gene expression counts were obtained after bulk RNA sequencing and used without normalization to day 0 values. Gene Set Variation Analysis (GSVA) was then performed with BTM gene sets to calculate normalized enrichment score (NES) values displayed on the heatmaps. BTM families are indicated by the color codes (left side), and BTM identifiers are listed on the right side.
Figure 9—figure supplement 4
Kinetics of average normalized enrichment score (NES) of blood transcription module (BTM) families after Armenia 2008 challenge.
NES values for BTM expression from individual animals were calculated by Gene Set Variation Analysis (GSVA) using non-normalized read counts. At each time point, NES values of BTMs belonging to one family were used to determine average family scores. Plots demonstrate average NES values for each family related to either innate (left column) or adaptive immunity (right column). At 4 and 7 days post-challenge (dpc), n=5 pigs in farm and specific pathogen-free (SPF) groups, and at 11 dpc, n=2 in farm group and n=4 in SPF group. Differences between protected and non-protected farm pigs were analyzed by unpaired t-test with Holm-Šídák’s correction for multiple comparisons; §p<0.05.
Figure 10
Model of protective immune responses against African swine fever virus (ASFV) challenge in immunized pigs.
The findings from blood transcription module (BTM) induction, systemic cytokine, and T-cell response analyses are integrated. Following immunization with the attenuated virus, an early IFN-α response activates antigen-presenting cells (APCs), triggering sustained cell cycle activation and clonal expansion of antigen-specific lymphocytes. This process generates plasma cells and effector T cells, with efficient memory T helper cell development emerging as a key correlate of protection. Upon challenge, a controlled IFN-α response and a rapid activation of cell cycle genes correlate with protection. The coordinated action of antibody-producing plasma cells and activated helper and cytotoxic T cells effectively controls viral replication while preventing a cytokine storm, ultimately conferring protection against ASFV. Figure was created with BioRender.com.
Tables
Table 1
List of antibodies used for flow cytometry.
| Antigen | Clone | Species/isotype | Fluorochrome | Source | Cat No. |
|---|---|---|---|---|---|
| CD3 | PPT3 | Mouse IgG1 | – | Hybridomas | – |
| CD4 | 74-12-4 | Mouse IgG2b | – | Hybridomas | – |
| CD8α | 76-2-11 | Mouse IgG2a | PE | BD Pharmingen | 559584 |
| CD8β | PG164A | Mouse IgG2a | – | WSU | PG164A |
| δ-TCR* | PGBL22A | Mouse IgG1 | – | WSU | PGBL22A |
| IFN-γ | P2G10 | Mouse IgG1 | PerCP-Cy5.5 | BD Pharmingen | 561481 |
| TNF-α | MAb11 | Mouse IgG1 | AF647 | BioLegend | 502916 |
| IgG1 | Polyclonal | Goat IgG | APC/CY7 | SouthernBiotech | 1070–19 |
| IgG2b | Polyclonal | Goat IgG | AF488 | Invitrogen | A-21141 |
| IgG2a | Polyclonal | Goat IgG | PE/Cy7 | Abcam | ab130787 |
-
*
The antibody against δ-TCR chain was coupled to biotin using Zenon Mouse IgG1 labeling kit (Z25052, Invitrogen, USA). Streptavidin coupled with BV421 was used as a conjugate (563259, BD Horizon, USA).
Additional files
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Correlates of protection against African swine fever virus identified by a systems immunology approach
eLife 14:RP107579.
https://doi.org/10.7554/eLife.107579.3
