![]() Instead of running t-tests on 22,000+ genes, we can fit a linear model to the data. Suppose we want to find genes that change expression between the cases that have distant metastases in under 5 years from those that don’t. # alternative hypothesis: true difference in means is not equal to 0 You already learned a tool to test if two groups have the same mean or a different mean. It seems that there is some separation between the groups, but since there is some variance, we can’t tell if the group means are different or not. Stripchart(exprs(mainz) ~ mainz.fac, method="jitter") Mainz.fac <- factor(mainz.class, labels= c("CLEAR","DM")) mainz.class <- (pData(mainz)$e.dmfs & pData(mainz)$t.dmfs < 5*365) We can then examine the expression of the selected gene in these groups. Let’s divide the arrays into two groups, patients that developed distant metastases within 5 years of diagnosis and those that did not. I’ll pick a row from the 22,000 probes in the arrays and plot the expression in each array for that gene. Library(hgu133a.db) # Loading required package: AnnotationDbi # Loading required package: stats4 # Loading required package: IRanges # Loading required package: S4Vectors # Loading required package: org.Hs.eg.db # Loading required package: DBI # data(mainz) # Attaching package: 'limma' # The following object is masked from 'package:BiocGenerics': # 'citation("Biobase")', and for packages 'citation("pkgname")'. # Vignettes contain introductory material view with # setdiff, sort, table, tapply, union, unique, unlist, unsplit # Welcome to Bioconductor # pmin.int, Position, rank, rbind, Reduce, rownames, sapply, # mapply, match, mget, order, paste, pmax, pmax.int, pmin, # grep, grepl, intersect, is.unsorted, lapply, lengths, Map, # colnames, do.call, duplicated, eval, evalq, Filter, Find, get, # anyDuplicated, append, as.ame, as.vector, cbind, # IQR, mad, xtabs # The following objects are masked from 'package:base': # parLapplyLB, parRapply, parSapply, parSapplyLB # The following objects are masked from 'package:stats': # clusterExport, clusterMap, parApply, parCapply, parLapply, # clusterApply, clusterApplyLB, clusterCall, clusterEvalQ, # Attaching package: 'BiocGenerics' # The following objects are masked from 'package:parallel': library(Biobase) # Loading required package: BiocGenerics # Loading required package: parallel # R-Studio will keep the packages loaded into a workspace, even if you exit and restart the program, as long as you save your workspace. ![]() Once packages have been installed, you need to load the package into each session that will use them. ![]()
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |