Activity

  • Glenn Magnussen posted an update 6 months, 1 week ago

    [23], has the benefit of realism. Alternatively, we could order Betulin impose a constant quantity of people per deme. (i) First, we could decide on a dividing person in the entire metapopulation with probability proportional to its fitness, and simultaneously suppress a different individual, chosen at random within the similar deme. Even so, in this case, men and women in demes of higher fitness would exhibit shorter lifespans, which is not realistic and may well introduce a bias. A second possibility could be to opt for a dividing person (in line with fitness) in each on the demes, and to simultaneously suppress a further person, chosen at random, in each deme. Nonetheless, within this case, unless migration events are far much less frequent than these collective division-death events (i.e., these D division-death events), the time interval between them becomes artificially discretized. This introduces biases unless the total migration rate mDN is significantly smaller sized than Nd, i.e. unless m d=D.Solutions 1 Simulation methodsOur simulations are based on a Gillespie algorithm [48,49] that we coded in the C language. Here we are going to describe our algorithm for the case of a metapopulation of D demes of identical size, that is the main scenario discussed in our operate. In our simulations, every deme features a fixed carrying capacity K e talk about this option additional in this section. 1.1 Algorithm. A number of different events occur in our simulations, each and every with an independent price: (ii)NN N NEach person divides at rate fg (1{Ni =K), where fg is the fitness associated with the genotype g [ f0,1,2g of the individual, and Ni is the current total number of individuals in the deme i [ ,D to which the individual belongs. This corresponds to logistic growth. If a dividing cell has gv2, upon division, its offspring (i.e., one of the two individuals resulting from the division) mutates with probability m, to have genotype gz1 instead of g. Each individual dies at rate d. Hence, at steady-state, Ni K(1{d=fi ), where fi is the average fitness of deme i. In practice, we choose d 0:1, and fitnesses of order one, thus Ni 0:9K. P Migration occurs at total rate m D Ni . Two different demes i 1 are chosen at random, an individual is chosen at random from each of these two demes, and the two individuals are exchanged. There is no geographic structure in our model, i.e. exchange between any two demes is equally likely.Consequently, while imposing a constant number of individuals is a good simulation approach for a non-subdivided population (see e.g. [28]), it tends to introduce biases in the study of metapopulations. While we chose to perform simulations with fixed carrying capacities in order to avoid any of these biases, we checked that, for small enough migration rates, our results are completely consistent with simulation scheme (ii) described above. This consistency check also demonstrates that it is legitimate to compare our simulation results obtained with fixed carrying capacities to our analytical work carried out with constant population size per deme.2 Crossing time of the champion demeIn this section, we give more details on the calculation of the average valley or plateau crossing time tc by the champion deme amongst D independent ones.