Welcome to Wei-Chien's Mosquito House
Abstract > Effect of Multiple Blood Meal Probabilities

Effect of Multiple Blood Meal Probabilities



How does the resulting blood meal probabilities depend on the given blood meal probabilities?
From laboratory data, it is known that the majority of mosquitoes only feed from two different human individuals at most. In fact, the estimated one blood meal probability is about 80% and 20 % for two blood meals. Therefore, in the first part of the experiment, we mostly focus on these two probabilities.

In order to test the effect of given blood-meal probabilities on the resulting multiple blood-meal probabilities. Eight sets of theoretical probabilities were tested to see how the resulting lower bounds of blood meal proportions depart from their theoretical numbers.

From laboratory data, we realized the majority of mosquitoes have only bitten two different human individuals at most. In fact, the estimated lower bound is about 80% one blood meal and 20 % two blood meals. Therefore, in the first part of the experiment, we mostly focus on these two given probabilities.

Graph 1: Comparison of lower bound proportions and their theoretical numbers. In this experiment, eight groups of probabilities are tested to compare, which are 90% one blood meal and 10% two blood meals, 85 % one blood meal and 15 %two blood meals, 80 % one blood meal and 20 %two blood meals … and so on. Human population = 1000. Mosquito population=324. The red line indicates theoretical numbers.




As Graph 1 shows, when four different loci are applied for analysis, the resulting single blood meal probabilities and resulting two blood meal probabilities follow the change of their theoretical numbers.

How many loci must be analyzed in order for the lower bound method to accurately predict the multiple meal probabilities? There are four loci to be analyzed, which are TPOX, PK211, D13, and CSF respectively. Are four loci sufficient to accurately estimate? If not, how many loci are needed? What if we calculate the multiple blood meal probabilities by only performing PCR on three, two or even one locus?