Function：

Differential metabolites analysis tool is used to screen out differential metabolites between two groups of samples. Multivariate statistical analysis is widely used in metabolomics due to the complexity of metabolomics datasets, which reveals the metabolic profiles of two samples and discovers differential metabolites. In addition, univariate analysis such as T-test can be also used in differential metabolites analysis.

Input：

Data format: Data format should be tab delimited. If your data is not TAB delimited, you can use Excel to convert it

1. Metabolites abundance table file

Input a tab delimited text file with headers. The first row is sample names and the first column is metabolite IDs. Values in the table are abundance of each metabolite.

2. Grouping file

Input a tab delimited text file where sample names in the first column and group names in the second column. Note that this grouping file is also necessary for experiments without biological replicates. We recommend biological replicates in metabolomics study.

3. Comparison file

Input a tab delimited text file where a row represents a comparison with the first column versus the second column.

Parameters：

1. Method for differential analysis：We offer three methods for differential analysis: T-test, Partial Least-Squares Discriminant Analysis (PLS-DA), and Orthogonal Partial Least-Squares Discriminant Analysis (OPLS-DA). You can choose one of these methods for differential metabolites analysis. PLS-DA is a supervised dimensionality reduction method, which groups data according to the experiment before dimensionality reduction, and is often used to distinguish metabolic profiles of two groups of samples and screen out differential metabolites. OPLS-DA incorporates an Orthogonal Signal Correction (OSC) filter into a PLS model, effectively separating Y-predictive variation from Y-uncorrelated variation in X.

2. P value：The test value of t-test for single variable statistic analysis,At the same time, multivariate statistical analysis and single variable statistical analysis were adopted to select the differential metabolites.Usually,we choose P-value＜0.05

3. VIP threshold：The Variable Importance in Projection (VIP), which reflects both the loading weights for each component and the variability of the response explained by this component, can be used for feature selection. VIP≥1 is often used to screen out differential metabolites. Note: Only when choosing PLS-DA or OPLS-DA method can this value be set.

4. Threshold of fold change：The expression fold change of metabolites between two groups. Normally, the threshold of fold change is 2.

Output：

A example for choose the method of PLS-DA：

1. diff_stat.xls：statistical table of differential metabolites number

2. diff_stat.png：differential metabolites histogram between comparison groups (scalar figure)

3. diff_stat.pdf：differential metabolites histogram between comparison groups (vector figure)

4. A-vs-B.all.xls：differential metabolites table

5. A-vs-B.filter.xls：significantly differential metabolites table

6. A-vs-B.pca.png：PCA score plot (scalar figure)

7. A-vs-B.pca.pdf：PCA score plot (vector figure)

8. A-vs-B.pc.xls：principal component table of PCA

9. A-vs-B.plsda.png：OPLS-DA score plot (scalar figure)

10. A-vs-B.plsda.pdf：OPLS-DA score plot (vec figure)

11. A-vs-B.permutation.png：The sorting test of the statistical model of OPLS-DA（the scalar figure）

12. A-vs-B.permutation.pdf：The sorting test of the statistical model of OPLS-DA（the vector diagram）

13. A-vs-B.tmp.Ttest.xls：Single variable statistical analysis results table

14. A-vs-B.vip.xls：The VIP table of each variable in the multivariate statistical model

##### Example： Metabolite abundance table file Comparison file Grouping file

### Output：

1、 differential metabolites histogram between comparison groups：

2、PCA score plot：

3、 OPLS-DA score plot：

3、 OPLS-DA statistical model permutation test proof diagram：