Bioinfor Skill: Analysis of Large Expression Data Sets
Once you starting trying to compare more than two experimental groups, things start to get very complicated. Well thought-out statistical methods are available for comparing two experimental conditions, but when you get the three or more all bets are off! As a result, you may want to turn to other analysis methods such as those describe below. These methods work well with both RNA-Seq and Microarray data, and can also be used with other data types, including ChIP-Seq, MethylC-Seq, or anything else! All of the methods here assume you have your data available as a normalized gene expression matrix, defined as a tab-delimited text file with each row corresponding to a gene and each column a different experiment (Doesn't need to be gene expression, but you need a 'matrix' of data). You may need to use Excel to format your data like this. When analyzing gene expression, it's best to use a common gene accession number or gene symbol for the Gene I...