Must-Read Tips:
1.Please ensure files are in the tab-delimited text format
(.txt). The abundance file should not have any row
with all zero values.
2.Rows represent samples and columns represent feature/metadata
information. Metadata file should at least contain the columns
of “Group” and “Cohort”. For training/test dataset, we support
binary classification metadata of the Group column. For
multi-class data, we recommend converting it into multiple
binary classifications for analysis. Corresponding feature count
and metadata files should align with the same row names.
3.The training and test datasets should share a minimum of 50%
common features for optimal performance.
4.To maintain consistency across all analyses, when uploading
the three sets of datasets (training, test, and non-target),
please ensure that the control group name remains uniform across
them.
Please refer to our sample datasets for the
appropriate data format
in the Try sample data Page.