get_cclasso_info: Get graph and networks using cclasso to measure correlation...
In ConesaLab/MDM: Complete MDM Network Analysis for 16S rRNA Data

Description Usage Arguments Value Examples

Measure taxa co-occurrence relationships using cclasso method and construct networks of taxa with significant direct co-occurrence relationships by this method.An igraph object to visualize networks or calculate network measures is returned as output. Class network and hub network plots are also plotted and returned as output.

1	get_cclasso_info(orig_phylo, met_name)

`orig_phylo`	Phyloseq including all taxa meeting threshold, produced from get_back_res_meeting_min_occ function
`met_name`	Name of environment

Returns List of 2 objects: 1. cclasso igraph. 2. List of 2 plots: Hub and Class network plots

`cclasso_igraph`	Igraph of cclasso network
`Plot_list`	List of 2 plots: 1. Class network plot, 2. Hub network plot

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (orig_phylo, met_name)
{
    hy_cclasso_m <- as.matrix(t(otu_table(orig_phylo)))
    hy_test_cclasso_f <- cclasso(hy_cclasso_m, counts = TRUE)
    hy_cclasso_corr_m <- as.matrix(hy_test_cclasso_f$cor_w)
    rownames(hy_cclasso_corr_m) = colnames(hy_cclasso_corr_m) = rownames(otu_table(orig_phylo))
    hy_cclasso_pval_m <- as.matrix(hy_test_cclasso_f$p_vals)
    dim(hy_cclasso_pval_m)
    rownames(hy_cclasso_pval_m) = colnames(hy_cclasso_pval_m) = rownames(otu_table(orig_phylo))
    hy_cclasso_cormat1 <- ifelse(hy_cclasso_pval_m <= 0.05, hy_cclasso_corr_m,
        0)
    diag(hy_cclasso_cormat1) <- 0
    hy_cclasso_cormat1_corr_thresh_0.6 <- abs(hy_cclasso_cormat1) >=
        0.6
    hy_cclasso_cormat1_adjmat <- Matrix::Matrix(hy_cclasso_cormat1_corr_thresh_0.6,
        sparse = TRUE)
    library(SpiecEasi)
    hy_cclasso_igraph <- igraph::adj2igraph(hy_cclasso_cormat1_adjmat,
        vertex.attr = list(name = taxa_names(orig_phylo)))
    met_cclasso_graph = hy_cclasso_igraph
    plot_a <- phyloseq::plot_network(hy_cclasso_igraph, orig_phylo,
        type = "taxa", color = "Class", label = NULL, title = paste(met_name,
            "CCLASSO network", sep = " ")) + theme(legend.position = "none")
    plot_b <- phyloseq::plot_network(hy_cclasso_igraph, orig_phylo,
        type = "taxa", color = "Genus", label = NULL, title = paste(met_name,
            "CCLASSO hub network", sep = " "), point_size = hub_score(hy_cclasso_igraph)$vector *
            10) + theme(legend.position = "none")
    all_plots <- list(plot_a, plot_b)
    all_info <- list(met_cclasso_graph, all_plots)
    return(all_info)
  }