LDA Score positive/negative

[fholz]

Dear all,

I have a question regarding results of LEfSe analysis for metagenomics data.
In my data “Streptococcus” is more abundant in No BI group vs. BI group and the LDA Score is shown as 3.5 (see screenshots attached)

Is it possible to reverse the direction of analysis in order to have a negative LDA Score for Streptococcus?

The reason is that in previous analyses comparison was the other way around and bacteria that were more abundant in No BI had negative LDA scores.

Thank you very much!

PS: When switched the order of the two groups in the metadata table, LDA score still was positive (see screenshots attached)

[jeff.xia]

The dot plot uses mini-heatmap for direction. You should use bar plot to show negative values

[fholz]

Thanks your for your response.

I chose Bar Plot. However, Streptococcus is still assigned with positive LDA Score.

[jeff.xia]

The negative or positive is mainly based on which is reference. You can try to use different labels (computers use alphabet order)

[Yao]

Hello, we have fix this issue. The result should be consistent between the boxplot and the dot/bar plot. You can try agin.
You can either edit the order of the groups before upload or use our metadata editor to change the order. The final result will be related to the order you have set.

Please let us know if you have more questions.