Commit c0d07104 authored by Samuel BERTRAND's avatar Samuel BERTRAND 🐉

Update README.md

parent 0744bfed
......@@ -176,6 +176,106 @@ Again two result tables a provided. One corresponding to the selected features c
The provided data corresponds to the “SSRankProduct” value, which corresponds to the normalized SSRankProduct. Thus the SSRankProduct values is used to select features of interest and can be modified within the function argument “lim.Pochemon”.
<h2>Comparison of the provided results</h2>
The different strategy provided rather orthogonal feature selection. Therefore, it might be important to test them all in order to highlight all induced features.
The function provides comparison of feature selection as follow, using method = “All”:
` result<-Coculture.analysis(data=data, SampleNames = rowheader , monocultureSamples = c("Fungi1","Fungi2"), cocultureSamples = "Fungi1VSFungi2", scaleC='standard', log10L=TRUE, Method = "All")`
Using such an approach, all strategies are applied and grouped together to provide a general list of all results containing various tables corresponding to: “result$Ttest”, “result$Wtest”, “result$OPLS.DA” and “result$Pochemon”. In addition, the “result$featuresSelection” contains all selected features considering all data analysis methods. This this latter table has the following structure:
<table>
<tr>
<td></td>
<th>var<sub>1</sub></th>
<th>var<sub>2</sub></th>
<th>...</th>
<th>var<sub>i</sub></th>
<th>...</th>
</tr>
<tr>
<th>Ttest.FoldChange</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>Ttest.p-value</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>Wtest.FoldChange</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>Wtest.p-value</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>OPLS.DA.VIP1 (pure metabolome1 / mixed metabolome)</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>OPLS.DA.VIP1 (pure metabolome2 / mixed metabolome)</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>OPLS.DA.p1 (pure metabolome1 / mixed metabolome)</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>OPLS.DA.p2 (pure metabolome2 / mixed metabolome)</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>OPLS.DA.p1+p2</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
<tr>
<th>SSRankProduct</th>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td></tr>
</table>
In addition, the function provides a box plot representing variable score distribution according to each data analysis methods.
<img src='https://gitlab.univ-nantes.fr/bertrand-s-1/pochermon/raw/master/Pictures/distribution.png' />
Finally, all feature selection is compared using a Venn Diagram highlighting common and specific feature selection.
<img src='https://gitlab.univ-nantes.fr/bertrand-s-1/pochermon/raw/master/Pictures/Venndiag.png' />
<h2>Further use</h2>
Such Pochemon approach is also compatible with variance decomposition through ANOVA, as reported in Geurts, B.P., Neerincx, A.H., Bertrand, S., Leemans, M.A.A.P., Postma, G.J., Wolfender, J.L., Cristescu, S.M., Buydens, L.M.C., Jansen, J.J., 2017. Combining ANOVA-PCA with POCHEMON to analyse micro-organism development in a polymicrobial environment. Anal. Chim. Acta 963, 1-16, http://doi.org/ 10.1016/j.aca.2017.01.064
<br />
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment