Name Mode Size
.github 040000
R 040000
data 040000
inst 040000
man 040000
src 040000
tests 040000
vignettes 040000
.Rbuildignore 100644 0 kb
.gitignore 100644 0 kb
DESCRIPTION 100644 2 kb
LICENSE 100644 1 kb
NAMESPACE 100644 5 kb
NEWS 100644 10 kb
README.Rmd 100644 4 kb
README.md 100644 5 kb
README.md
sitePath: phylogeny-based sequence clustering using site polymorphism ================ The below demonstrates the result of phylogeny-based sequence clustering for a H3N2 virus dataset (included in the package) ``` r library(sitePath) ``` ## Warning in register(): Can't find generic `scale_type` in package ggplot2 to ## register S3 method. ``` r data(h3n2_align) # load the H3N2 sequences data(h3n2_tree) # load the corresponding phylogenetic tree options(list("cl.cores" = 10)) # Use 10 cores for multiprocessing paths <- lineagePath(h3n2_tree, alignment = h3n2_align, Nmin = 0.05) ``` ## The "tree" object is not bifurcated and resolved by "multi2di" function. ## Using 10 cores.. ## Multiprocessing ended. ``` r minEntropy <- sitesMinEntropy(paths) ``` ## Using 10 cores.. ## Multiprocessing ended. ``` r p1 <- plotSingleSite(paths, site = 208) # The site polymorphism of site 208 on the tree p2 <- plotSingleSite(minEntropy, site = 208) # The result of clustering using site 208 gridExtra::grid.arrange(p1, p2, ncol = 2) ``` ![](inst/example-1.png)<!-- --> ``` r grp1 <- extractTips(paths, 208) # Grouping result using site polymorphism only grp2 <- extractTips(minEntropy, 208) # Phylogeny-based clustering result ``` # Installation [R programming language](https://cran.r-project.org/) \>= 4.1.0 is required to use `sitePath`. The stable release is available on [Bioconductor](https://bioconductor.org/packages/sitePath/). ``` r if (!requireNamespace("BiocManager", quietly = TRUE)) install.packages("BiocManager") BiocManager::install("sitePath") ``` The installation from [GitHub](https://github.com/wuaipinglab/sitePath/) is in experimental stage but gives the newest feature: ``` r if (!requireNamespace("remotes", quietly = TRUE)) install.packages("remotes") remotes::install_github("wuaipinglab/sitePath") ``` # QuickStart The following is a quick tutorial on how to use `sitePath` to find fixation and parallel sites including how to import data, run analysis and visualization of the results. ## 1. Data preparation You need a *tree* and a *MSA* (multiple sequence alignment) file and the sequence names have to be matched! ``` r library(sitePath) # Load the sitePath package # The path to your tree and MSA files tree_file <- system.file("extdata", "ZIKV.newick", package = "sitePath") alignment_file <- system.file("extdata", "ZIKV.fasta", package = "sitePath") tree <- read.tree(tree_file) # Read the tree file into R align <- read.alignment(alignment_file, format = "fasta") # Read the MSA file into R ``` ## 2. Run analysis `Nmin` and `minSNP` are the respective parameters for finding fixation and parallel sites (18 and 1 are used as an example for this dataset). The default values will be used if you don’t specify them. ``` r options(list("cl.cores" = 1)) # Set this bigger than 1 to use multiprocessing paraFix <- paraFixSites(tree, alignment = align, Nmin = 18, minSNP = 1) # Run analysis to find fixation and parallel sites paraFix ``` ## This is a 'paraFixSites' object ## ## fixation sites: ## 139, 894, 2074, 2086, 2634, 3045, 988, 1143, 2842, 3398, 107, 1118, 3353 ## ## parallel sites: ## 105, 2292, 1264, 918, 1226, 1717, 988, 2611, 2787, 2749, 3328, 3162, 1857, 2445, 358, 1404, 3046, 791, 1180, 1016, 1171, 1327, 3076, 106, 2357, 916, 1303, 969, 573, 2909, 2122, 940 ## ## paraFix sites: ## 988 ## 3. Fixation sites Use `allSitesName` and set `type` as “fixation” to retrieve fixation sites name ``` r allSitesName(paraFix, type = "fixation") ``` ## [1] "139" "894" "2074" "2086" "2634" "3045" "988" "1143" "2842" "3398" ## [11] "107" "1118" "3353" Use `plotFixationSites` to view fixation sites ``` r plotFixationSites(paraFix) # View all fixation sites on the tree ``` ![](inst/plot_fixSites-1.png)<!-- --> ``` r plotFixationSites(paraFix, site = 139) # View a single site ``` ![](inst/plot_fixSites-2.png)<!-- --> ## 4. Parallel sites Use `allSitesName` and set `type` as “parallel” to retrieve parallel sites name ``` r allSitesName(paraFix, type = "parallel") ``` ## [1] "105" "2292" "1264" "918" "1226" "1717" "988" "2611" "2787" "2749" ## [11] "3328" "3162" "1857" "2445" "358" "1404" "3046" "791" "1180" "1016" ## [21] "1171" "1327" "3076" "106" "2357" "916" "1303" "969" "573" "2909" ## [31] "2122" "940" Use `plotParallelSites` to view parallel sites ``` r plotParallelSites(paraFix) # View all parallel sites on the tree ``` ![](inst/unnamed-chunk-1-1.png)<!-- --> ``` r plotParallelSites(paraFix, site = 105) # View a single site ``` ![](inst/unnamed-chunk-1-2.png)<!-- --> # Read more The above uses wrapper functions but the analysis can be dissembled into step functions (so you can view the result of each step and modify parameters). Click [here](https://wuaipinglab.github.io/sitePath/articles/sitePath.html) for a detailed breakdown of the functionality. # Getting help Post on Bioconductor [support site](https://support.bioconductor.org/) if having trouble using `sitePath`. Or open an [issue](https://github.com/wuaipinglab/sitePath/issues/new?assignees=&labels=&template=bug_report.md&title=) if a bug is found.