# Median-Joining Network This function computes the median-joining network (MJN) as described by Bandelt et al. (1999). ```r mjn(x, epsilon = 0, max.n.cost = 10000, prefix = "median.vector_", quiet = FALSE) ## S3 method for class 'mjn' plot(x, shape = c("circles", "diamonds"), bg = c("green", "slategrey"), labels = FALSE, ...) ``` ## Arguments - `x`: a matrix (or data frame) of DNA sequences or binary 0/1 data; an object of class `"mjn"` for `plot`. - `epsilon`: tolerance parameter. - `max.n.cost`: the maximum number of costs to be computed. - `prefix`: the prefix used to label the median vectors. - `quiet`: a logical value; by default, the progress of the calculatins is printed. - `shape, bg`: the default shapes and colours for observed haplotypes and median vectors. - `labels`: by default, the labels of the haplotypes are printed. - ``...``: other arguments passed to `plot.haploNet`. ## Details MJN is a network method where unobserved sequences (the median vectors) are reconstructed and included in the final network. Unlike `mst`, `rmst`, and `msn`, `mjn` works with the original sequences, the distances being calculated internally using a Hamming distance method (with `dist(x, "manhattan")` for binary data or `dist.dna(x, "N")` for DNA sequences). The parameter `epsilon` controls how the search for new median vectors is performed: the larger this parameter, the wider the search (see the example with binary data). If the sequences are very divergent, the search for new median vectors can take a very long time. The argument `max.n.cost` controls how many such vectors are added to the network (the default value should avoid the function to run endlessly). The arguments `shape` and `bg` must be of length two (unlike in `plot.haploNet`). It is possible to have more flexibility when plotting the MJN by changing its class, for instance with the output in the examples below: `class(nt0) <- "haplotNet"`. ## Note Since `pegas` 1.0, `mjn` is expected to run in reasonable times (less than 15 sec with 100 sequences). Bandelt et al. (1999) reported long computing times because of the need to compute a lot of median vectors. Running times also depend on the level of polymorphism in the data (see above). ## Returns an object of class `c("mjn", "haploNet")` with an extra attribute (data) containing the original data together with the median vectors. ## References Bandelt, H. J., Forster, P. and Rohl, A. (1999) Median-joining networks for inferring intraspecific phylogenies. **Molecular Biology and Evolution**, 16 , 37--48. ## Author(s) Emmanuel Paradis ## See Also `haploNet`, `mst` ## Examples ```r ## data in Table 1 of Bandelt et al. (1999): x <- c(0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 0, 1, 1) x <- matrix(x, 4, 9, byrow = TRUE) rownames(x) <- LETTERS[1:4] (nt0 <- mjn(x)) (nt1 <- mjn(x, 1)) (nt2 <- mjn(x, 2)) plot(nt0) ## Not run: ## same like in Fig. 4 of Bandelt et al. (1999): plotNetMDS(nt2, dist(attr(nt2, "data"), "manhattan"), 3) ## End(Not run) ## data in Table 2 of Bandelt et al. (1999): z <- list(c("g", "a", "a", "a", "a", "a", "a", "a", "a", "a", "a", "a"), c("a", "g", "g", "a", "a", "a", "a", "a", "a", "a", "a", "a"), c("a", "a", "a", "g", "a", "a", "a", "a", "a", "a", "g", "g"), c("a", "a", "a", "a", "g", "g", "a", "a", "a", "a", "g", "g"), c("a", "a", "a", "a", "a", "a", "a", "a", "g", "g", "c", "c"), c("a", "a", "a", "a", "a", "a", "g", "g", "g", "g", "a", "a")) names(z) <- c("A1", "A2", "B1", "B2", "C", "D") z <- as.matrix(as.DNAbin(z)) (ntz <- mjn(z, 2)) ## Not run: ## same like in Fig. 5 of Bandelt et al. (1999): plotNetMDS(ntz, dist.dna(attr(ntz, "data"), "N"), 3) ## End(Not run) ```

mjn function