I have a question regarding interpretation of alpha diversity boxplots (Shannon, Bray-Curtis, etc.). How are the whisker and box lengths determined? Does the box represent the upper and lower quartile with the whiskers showing the highest and lowest observation? If that is the case, do points/circles outside of the whiskers represent outliers to the data? I can’t seem to find a clear answer specific to QIIME anywhere.
Thank you in advance!
Hello Ahmad,
This link:
Article Alterations in Gastric Microbiota After H. Pylori Eradicatio...
describes the box plots as good old Tukey box-and-whisker plot rules (e.g., whiskers extend up to 1.5X mid-band spread).
Good luck with your work.
Hi Ahmad Ali ,
If I have understood correctly, what you need to know is an image of a boxplot and the interpretation of data distribution in it. Almost in all boxplots (which are non-parametric) including the those with Shannon index, there are 4 quartiles (Q1-4) in which the median is in the middle and whiskers extend to each extremums (Max and min of your data).
Most (depending on the CI) of your data are distributed in the Interquartile Range (IQR; normally 25-75 percentile) which is the box itself (Q1-3). The length of the IQR or the box is determined by the frequency and statistical distribution of your data if they are arranged in a way with median in the middle.
Actually according to their very names, outliers are data that are significantly different from your normally distributed data. They can even show skewness or the distribution or statistical error as well. Thus they are often out of the range or max and min (whiskers). There are different ways to detect outliers (Peirce, Tukey, ...). So the answer to all of your questions: Yes.
Hope this helps.
Boxplot, stem-and-leaf and histogram are the three types of graphical representation useful in analyzing distributions. You can select any of the three types of graphical representations because they are similar, but avoiding their simultaneous use in a research document, as well as representing a single variable.
Even if some representations are more complex, I recommend using only the graphic representation you know well.
The boxplot representation includes 50% of the distribution values, the vertical lines being drawn to a length of 1.5x H, where H is the distance between the box boundaries (interquartile deviation Q3-Q1).
The points marked beyond the horizontal limits of the representation mark the marginal values of the distribution (outliers) and the extreme values.
Best regards,
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology