Shulou(Shulou.com)06/01 Report--

How to use DESeq2 for PCA and cluster analysis of expression, I believe that many inexperienced people do not know what to do. Therefore, this paper summarizes the causes and solutions of the problem. Through this article, I hope you can solve this problem.

After the expression of gene / transcript is obtained, the relationship between biological samples and experimental design is usually examined and analyzed by the following three types of charts.

1. Clustering tree of samples

The expression data of all samples are used to cluster the samples. In theory, if there is no problem with the sample and the experimental operation, then the biological repetitive samples belonging to the same group will come together. The schematic diagram is as follows

In the above picture, the name of the sample is replaced by the group, so you can see that the samples with the same condition are grouped together.

2. PCA diagram

The dimensionality reduction is carried out by principal component analysis, and the distribution of sample points and the location of base points are displayed on a two-dimensional or three-dimensional plane, and we can also see whether the samples belonging to the same group are together and whether the samples between different groups are obviously separated, as shown below.

It can be seen from the figure that the samples with different conditions are clearly distinguished, while the biological repetition is relatively close, indicating that the consistency of biological repetition and the difference of different groups are good.

3. Heat map

Compared with the sample cluster tree, the heat map contains more information, for example, it can directly show the difference of expression between different groups, and it is also one of the common visualization methods, as shown below.

As long as there is a sample expression matrix, DESeq2 can easily draw the above three kinds of charts. But should we choose the original expression matrix or the normalized expression matrix to draw? Or are there any other options?

If the input matrix is different, the conclusion will be different. Because there are some differences in the level of gene expression among different samples, the effect is not ideal whether using the original or normalized expression matrix. In order to solve this problem, DESeq2 proposed two conversion algorithms of count value, rlog and VST conversion.

1. Rlog conversion

The use of the rlog transformation is as follows

Rld

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.