Data from: Host genetics and the skin microbiome independently predict parasite resistance
Data files
Feb 07, 2026 version files 521.63 KB
-
Discrete_abundances.csv
45.18 KB
-
Kramp_et_al_Micosatellite_data.csv
14.04 KB
-
ParentalSelection.csv
143.32 KB
-
README.md
4.95 KB
-
SelectionData2.csv
281.58 KB
-
Tableofblanks.csv
1.34 KB
-
TaxaDifferencesandabundances.csv
18.80 KB
-
weighted_distance-matrix.csv
12.43 KB
Abstract
Host responses to parasite infection involve several interacting systems. Host genetics determine much of the response, but it is increasingly clear that the host-associated microbiome also plays a role. Genetically determined systems and the microbiome can also interact; for example, the microbiome can modulate the immune response, and vice versa. However, it remains unclear how such interactions between the host immune system and the microbiome may influence the host’s overall response to parasites. To investigate how host genetics and the microbiome interact to shape responses to parasites, we imposed truncation selection on Trinidadian guppies (Poecilia reticulata) for low and high resistance to the specialist ectoparasite Gyrodactylus turnbulli. After 3-6 generations of breeding without parasites, we sampled the skin-associated microbiome and infected fish from each line. We applied Dirichlet Multinomial Modeling (DMM) machine-learning to identify bacterial community types across lines and evaluated how selection line and community type explained variations in infection severity. Our findings showed that among females, the resistant line had significantly lower infection severity, while the susceptible line had higher infection severity. Among males, only the susceptible line experienced higher infection severity compared to the other lines. Line did not explain skin microbial diversity, structure or composition. Our DMM analysis revealed three distinct bacterial community types, independent of artificial selection lines, which explained just as much variation in infection load as selection line. Overall, we found that the microbiome and host genetics independently predict infection severity, highlighting the microbiome's active role in host-parasite interactions.
Dataset DOI: 10.5061/dryad.zs7h44jmn
Description of the data and file structure
In this study, we wanted to understand how certain factors can affect the outcome of parasitic infections. These factors include artificial selection, the gender of the fish, and the community of microorganisms living on the fish's skin.
To set up the artificial selection process, we infected 200 Poecilia reticulata fish (half male and half female) with a parasite called Gyrodactylus turnbulli. We closely monitored the infection, then divided the fish into two groups: one group with the top 30 % of fish having the highest average worm burden (susceptible line) and the other group with the bottom 30 % of fish having the lowest average worm burden (resistance line). We also had control groups that were made up of uninfected fish randomly selected from the same population. Data on these fish is found in ParentalSelection.csv.
After this initial selection process, we allowed the fish from each group to breed within their respective lines for four to seven generations without exposure to the parasite. Before infecting the offspring from the artificial selection, we collected swabs from the adult fish to measure the communities of microorganisms living on their skin.
SelectionData2.csv: In our analysis, the most important data can be found in the file 'SelectionData2.csv,' which contains various variables used in our main statistical models:
- fishID: Identity of the focal fish
- line: The 30 % most intense infection founded the susceptible line and the 30 % least intense infections founded the resistance lines. The controls lines were founded with the same number of parents of randomly selected uninfected fish from the same population.
- JulienDays: Experimental dates when the fish were swabbed and infected
- sex: Sex of the fish
- length, weight: Length in mm and weight (mass) in mg of the fish
- respreL: The residuals of length of the focal fish on sex. Females tend to be larger than males: using the residuals of this relationship allows us to test for both size and sex differences in behavior.
- PREsmi The residuals of the scaled mass index of fish
- rSMIpre Body condition - (rescaled so that males and females could be compared) scaled mass index
- AUC the Area Under the Curve for worm days / time "infection integral"
This data was subsetting to run analysis appropriately, all dataframes are explained below.
The following sub-setting of Dataframe's explained:
IO Data frame: Only offspring that were infected with G.turnbulli, and will be used in GLMM model.
female and male Data frames: Subset F4-7 offspring fish that were infected with G. turnbulli parasite sub-setting to contain only female fish or male fish for post-hoc analysis
There is evidence that the tiny organisms living in and on hosts, called the microbiome, might be passed down from one generation to the next and have evolved alongside their hosts (a component of the "holobiont" theory). To find out whether the microbiome on the skin of our fish played a role in making them more or less resistant to infections, we collected samples from 51 fish: Control (n = 15), Resistant (n = 18), Susceptible (n = 17), Sham-infected fish (n = 20). We then analyzed the communities of microorganisms in these samples using 16s rRNA sequencing. Our goal was to see if the fish from different lines had different associated skin microbiomes, and if these microbiomes were related to how susceptible or resistant they were to infections.
In continuation of the above analysis, the most important data can be found in the file 'SelectionData.csv.' This section expands to microbial community dataframes:
- weighted_distance-matrix.csv: Beta-diversity obtained from Qiime2 export pipeline- Unweighted and Weighted Unifrac distance matrices for PCoA creation and PERMANOVA analysis to which factor explains microbial variation on guppies' skin
- Discrete_abundances.csv: Relative frequencies of bacterial taxa data table exported from Qiime2 after DADA2 processing and SILVA classification
Sub-setting of SelectionData.csv explains:
M1 Data frame: Subset (n = 51) F4-7 offspring fish (sham infected included) that were swabbed and sent for 16s rRNA 515F--806R sequencing
MicroDat Data frame: Subset (n = 31) of parasite-infected offspring fish that were swabbed and sent for 16s rRNA 515F--806R sequencing
Kramp et al Microsatellite data.xlsx: All microsatellite data for genotyping fish
TaxaDifferencesandabundances.csv: All taxa and count using DMM modeling
Tableofblanks.csv: All features found in DNA extraction blanks
Kramp_et_al_Micosatellite_data.csv: Contains information on Guppy Microsatellite Locus and genotypes.