Newer
Older
## Prevalence of BTV/EHDV in Culicoides species in La Reunion.
T Balenghien, C Cetre Sossah
Question by Thomas Balenghien (2021-01-08 and 2021-03-16)
Insects of different species are captured in the field at different sites on different dates and tested for the presence of viruses.
As we cannot test all the insects, __we test batches__ (pools), and generally a small proportion (e.g. 3 batches of 50 out of the 8,000 individuals collected).
From this we can __estimate a percentage of infection by species, date and site__.
The objective is to know if there are __differences in infection rates__ between species for example.
One could think of a statistical model with the species as a fixed effect and the dates and sites as random effects. The problem is that we have an __over-abundance of 0s__, and in fact very few positive batches, which complicates this type of approach.
Another issue is that pools are of considerably different sizes, ranging from pools of 1 single insect to 50.
Thus, the information provided by each pool on the underlying prevalence of the population depends strongly on its size.
## Reports
- [Exploratory data analysis](https://umr-astre.pages.mia.inra.fr/insect-infection-rates/descriptive.html)
- [Pooled-Bernoulli likelihood](https://umr-astre.pages.mia.inra.fr/insect-infection-rates/pooled-bernoulli.html)
- [Model building](https://umr-astre.pages.mia.inra.fr/insect-infection-rates/insect_infection_rates.html)
- [Results](https://umr-astre.pages.mia.inra.fr/insect-infection-rates/results.html)