Taxonomic classification revisited.

GeneCreek has developed a taxonomic classification algorithm for amplicon microbiome sequencing called cur|sor, which outperforms the algorithms integrated in the state-of-the-art public microbiome analysis pipelines such as Qiime and Mothur. With cur|sor, over 80% of human gut Illumina 16SV4 sequences, including singletons, are classified at the species level.

cursor

Exploratory data analysis.

Working with species-level resolution of the microbiome composition instead of family/genus level resolution is a game changer, yet downstream exploratory statistics need to fit the nature of the input data. Whereas discovery methods such as LEfSe continue to work well for abundant and relatively prevalent species, low abundant, less prevalent species actually represent themselves as semi-discrete independent variables. GeneCreek has implemented microbiome-wide Chi-squared and Tukey HSD testing methods which address these different situations.
Image

Co-occurrence and mutual exclusion.

Using meta-cohort analysis, GeneCreek is capable of predicting synergy or competitive exclusion between species. At present, we have over 6,000 cur|sor analyzed clinical gut samples for these predictions. No patient metadata is required for this analysis. One of the uses is the discovery of beneficial bacteria, typically GRAS/QPS* species, suitable for use in probiotics compositions, with antagonistic activity against particular pathogens or pathobionts - commensal bacteria with negative influence on health in a particular situation.

*generally recognized as safe (FDA denomination) or qualified presumption of safety (EFSA denomination)

Image