http://picrust.github.io/picrust/
PICRUSt (pronounced “pie crust”) is a bioinformatics software package designed to predict metagenome functional content from marker gene (e.g., 16S rRNA) surveys and full genomes.
PICRUSt estimated the functional potential of microbial communities given a marker gene survey and the set of currently sequenced reference genomes with an accuracy of 80-90% on human gut communities.
Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Langille, M. G.I.*; Zaneveld, J.*; Caporaso, J. G.; McDonald, D.; Knights, D.; a Reyes, J.; Clemente, J. C.; Burkepile, D. E.; Vega Thurber, R. L.; Knight, R.; Beiko, R. G.; and Huttenhower, C. Nature Biotechnology, 1-10. 8 2013.
Hi Shi,
Thanks for asking. As part of the PICRUSt package we have a script called compare_biom.py that can compare to OTU tables in BIOM format. The best way to test metagenome prediction accuracy would be to compare PICRUSt predictions of metagenomes with sequenced metagenomes from the same set of samples. As part of the PICRUSt paper in Nature Biotechnology (http://www.nature.com/nbt/journal/vaop/ncurrent/full/nbt.2676.html) we did this comparison for all of the Human Microbiome Project samples with paired 16S rRNA genes and shotgun metagenomes (see Fig 2 and Fig 3), although we didn't split accuracy out by body site. Pooled across all body sites, Spearman r for predicted vs. sequenced metagenomes was 0.82, and examination of the PCoA plots/gene abundance plots in Fig 2 suggest that oral sites aren't an outlier in terms of accuracy.
We also have a google group for PICRUSt-related help questions that is monitored by myself and Morgan more closely than ResearchGate. You can find that here: https://groups.google.com/forum/#!forum/picrust-users.
Best,
Jesse
Hi Shi,
Thanks for asking. As part of the PICRUSt package we have a script called compare_biom.py that can compare to OTU tables in BIOM format. The best way to test metagenome prediction accuracy would be to compare PICRUSt predictions of metagenomes with sequenced metagenomes from the same set of samples. As part of the PICRUSt paper in Nature Biotechnology (http://www.nature.com/nbt/journal/vaop/ncurrent/full/nbt.2676.html) we did this comparison for all of the Human Microbiome Project samples with paired 16S rRNA genes and shotgun metagenomes (see Fig 2 and Fig 3), although we didn't split accuracy out by body site. Pooled across all body sites, Spearman r for predicted vs. sequenced metagenomes was 0.82, and examination of the PCoA plots/gene abundance plots in Fig 2 suggest that oral sites aren't an outlier in terms of accuracy.
We also have a google group for PICRUSt-related help questions that is monitored by myself and Morgan more closely than ResearchGate. You can find that here: https://groups.google.com/forum/#!forum/picrust-users.
Best,
Jesse
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R