Using Two-Part Statistics to Analyze Microbial Sequence Data
Author Information
Author(s): Brandie D. Wagner, Charles E. Robertson, J. Kirk Harris
Primary Institution: University of Colorado Denver
Hypothesis
Can a two-part statistic effectively identify taxa that differ between two groups in microbial ecology studies?
Conclusion
The two-part statistic outperforms traditional statistical tests in analyzing sequence data from microbial ecology studies.
Supporting Evidence
- The two-part statistic was shown to handle the complexities of sequence count data better than traditional methods.
- In the cystic fibrosis study, 12 species level taxa showed significant differences using the two-part test.
- The environmental cenote dataset revealed 79 genus level taxa with significant differences using the two-part test.
Takeaway
This study shows a new way to analyze data about tiny living things, helping scientists understand which ones are different in sick and healthy people.
Methodology
The study used two datasets, one from cystic fibrosis patients and another from cenote samples, to compare the performance of a two-part statistic against t-tests and Wilcoxon tests.
Limitations
The study's sample sizes were relatively small, which may affect the generalizability of the results.
Participant Demographics
The clinical dataset included samples from cystic fibrosis patients and healthy controls.
Statistical Information
P-Value
<0.01
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website