# #——load-packages,包括= FALSE --------------------------------------------- 美元knitr: opts_chunk设置(图。宽度= 5,fig.height = 3) suppressPackageStartupMessages({库(dplyr)库(dplyr)图书馆(ggplot2)图书馆(tidyr)图书馆(forcats)图书馆(phenopath ) }) ## ---- 模拟数据 ------------------------------------------------------------ set.seed sim < - simulate_phenopath (123 l) () ## ---- sim-structure ------------------------------------------------------------ 打印(str (sim卡 )) ## ---- 一些基因,fig.width = 7 ------------------------------------------------ genes_to_extract < - c(1、3、11、13、21、23、31、33)expression_df < - as.data.frame (sim $ y [, genes_to_extract])名称(expression_df) < - paste0(“gene_”,genes_to_extract) df_gex < - as_tibble (expression_df) % > %变异(x =因素(sim [[x]]), z = sim [[' z ']]) % > %收集(基因表达、- x - z) ggplot (df_gex, aes (x = z, y =表达式,颜色= x)) + geom_point () + facet_wrap(~基因,nrow = 2) + scale_color_brewer(面板= " set2。中的 ") ## ---- 主成分分析,fig.show = '举行 '------------------------------------------------ pca_df < - as_tibble (as.data.frame (prcomp (sim $ y) $ x[1:2])) % > %变异(x =因素(sim [[x]]), z = sim [[' z ']]) ggplot (pca_df, aes (x = PC1, y = PC2,颜色= x)) + geom_point () + scale_colour_brewer(面板=“set2”中的)ggplot (pca_df, aes (x = PC1, y = PC2,颜色= z)) + geom_point () ## ---- 看到结果,缓存= TRUE -------------------------------------------------- 符合< - phenopath (sim $ y, sim $ x, elbo_tol e-6 = 1,瘦= 40)打印(适合)# #——plot-elbo ---------------------------------------------------------------- plot_elbo(适合)# #——plot-results fig.show =“持有”,fig.width = 2.5, fig.height = 2.5——qplot (sim美元z,轨迹(适合))+ xlab(“真正的z”)+ ylab (Phenopath z) qplot (sim美元z, pca_df PC1美元)+ xlab(“真正的z”)+ ylab(“PC1 ") ## ---- print-correlation -------------------------------------------------------- 软木(sim z,美元轨迹(适合 )) ## ---- beta-df fig.width = 6,fig.height = 3 ----------------------------------- gene_names < - paste0(“基因”,seq_len (ncol(适合m_beta美元)))df_beta < - data_frame(β= interaction_effects(适合),beta_sd = interaction_sds(适合),is_sig = significant_interactions(适合),基因= gene_names) df_beta基因< - fct_relevel美元(美元df_beta基因,gene_names) ggplot (df_beta, aes (x =基因,y =β,颜色= is_sig)) + geom_point () + geom_errorbar (aes (ymin =β- 2 * beta_sd ymax =β+ 2 * beta_sd)) +主题(axis.text。x = element_text(角= 90,hjust = 1), axis.title.x = element_blank ()) + ylab(表达式(β))+ scale_color_brewer(面板=”关于我校“name = "意义重大 ") ## ---- graph-largest-effect-size ------------------------------------------------ which_largest < - which.max (df_betaβ美元)df_large < - data_frame (y = sim [[y]] [, which_largest], x =因素(sim [[x]]), z = sim [[' z ']]) ggplot (df_large, aes (x = z, y = y,颜色= x)) + geom_point () + scale_color_brewer(面板=“set2”中的)+ stat_smooth () ## ---- construct-sceset,警告= FALSE ---------------------------------------- suppressPackageStartupMessages(图书馆(SummarizedExperiment)) exprs_mat < - t (sim $ y) pdata < - data.frame南加州爱迪生公司(x = sim $ x) <——SummarizedExperiment(化验= (exprs = exprs_mat)列表,colData = pdata) sce # #——example-using-expressionset eval = FALSE -------------------------------- # 符合< - phenopath(预计,sim $ x) # 1 #适合< - phenopath(,“x”)# 2 #适合<——phenopath (sce ~ x) # 3 # #——initialisation-examples eval = FALSE ------------------------------------ # 符合< - phenopath (sim $ y, sim $ x, z_init = 1) # 1,初始化第一主成分#适合<——phenopath (sim $ y, sim $ x, z_init = sim $ z) # 2,预置为真值#适合<——phenopath (sim $ y, sim $ x, z_init =“随机”)# 3,随机初始化# #——cavi-tuning,eval = FALSE ------------------------------------------------ # 符合< - phenopath (sim $ y, sim $ x, #麦克斯特= 1000,# 1000迭代马克斯•# elbo_tol = 1依照# ELBO < 0.02%时考虑模型融合变化#薄= 20 #计算ELBO每20迭代 # ) ## ---- sessioninfo -------------------------------------------------------------- sessionInfo ()