Choose an Analysis Method

Every method the server offers is listed below. Pick a method to open its dedicated configuration page. New to the platform? Try the Quick Analysis Wizard for a guided 3-step Random Forest setup.

Data Preparation

Preprocessing Help

Impute missing values, normalise features and handle outliers. Produces a derived dataset for downstream use.

Needs: One dataset with numeric columns.

Configure
Feature Selection Help

Rank and select the most informative features (importance, SelectKBest, RFE, SFS and more). Produces a derived dataset with the selected features plus the target.

Needs: One dataset with a numeric target column.

Configure
Feature Engineering Help

Create derived features — ratios, sums, products, means, log1p, polynomial interactions, longitudinal summaries, one-hot columns, or a wide→long reshape — saved as a new dataset for downstream analysis.

Needs: Numeric columns (plus a group column for one-hot).

Configure

Statistical Analysis

Correlation Analysis Help

Compute a pairwise correlation matrix (Pearson, Spearman or Kendall) across all numeric features.

Needs: One dataset with at least two numeric columns.

Configure
Univariate Screening Help

Pre-modeling screening: per-feature Pearson, Spearman and ANOVA F statistics against a target with FDR correction. Produces a derived dataset of significant features.

Needs: One dataset with a numeric target column.

Configure
t-Test Help

Compare the mean of a numeric variable between two groups (independent or paired samples).

Needs: A numeric value column and a group column with exactly 2 groups.

Configure
One-way ANOVA Help

Compare the mean of a numeric variable across two or more groups.

Needs: A numeric value column and a group column with 2+ groups.

Configure
Chi-Square Test Help

Test whether two categorical variables are independent using a contingency table of observed counts.

Needs: Two categorical columns.

Configure
Mann-Whitney U Test Help

Non-parametric comparison of a numeric variable between two groups (no normality assumption).

Needs: A numeric value column and a group column with exactly 2 groups.

Configure
Kruskal-Wallis Test Help

Non-parametric comparison of a numeric variable across two or more groups.

Needs: A numeric value column and a group column with 2+ groups.

Configure
Repeated-Measures ANOVA Help

Within-subject ANOVA over wide-format timepoint columns (e.g. JUN/JUL/AUG cover), with an optional between-subject factor. Saves the melted long table as a derived dataset.

Needs: Two or more numeric timepoint columns, one row per subject.

Configure
Count GLM (Poisson / NB) Help

Generalised linear model for count responses — Poisson or negative-binomial family with categorical predictors automatically dummy-coded.

Needs: A non-negative count column and one or more predictors.

Configure
Mixed-Effects Model (LMM) Help

Linear mixed model with a random intercept per group — handles nested/repeated designs such as plots within pastures or ranches.

Needs: A numeric response, a fixed effect, and a grouping column.

Configure
Clustering Help

Group plots or samples into clusters (K-Means or hierarchical) on z-scored numeric features. Saves a derived dataset with a cluster column for downstream use (e.g. management zones).

Needs: One dataset with at least one numeric column.

Configure
Ordination (PCA / PCoA / NMDS) Help

Reduce a samples-by-features table to a 2-D ordination: PCA, principal coordinates analysis (PCoA) or non-metric multidimensional scaling (NMDS) with a choice of distance metric.

Needs: One dataset with at least two numeric columns.

Configure

Community Ecology

Alpha Diversity Help

Per-sample diversity indices (richness, Shannon, Simpson, Pielou evenness) over a community matrix; optional group summaries. Saves indices as a derived dataset.

Needs: A samples-by-features community matrix.

Configure
Beta Diversity (PERMANOVA + ANOSIM) Help

Pairwise distance matrix (Bray-Curtis, Jaccard, ...) with PERMANOVA and ANOSIM group-composition tests plus a distance heatmap.

Needs: Community matrix plus a categorical group column.

Configure
Mantel Test Help

Correlates two distance matrices (e.g. community dissimilarity vs environmental distance) with a permutation p-value.

Needs: Two non-overlapping sets of numeric columns.

Configure
Indicator Species (IndVal) Help

Dufrene-Legendre indicator values: which features are specific to and frequent within each group.

Needs: Community matrix plus a categorical group column.

Configure
Morans I (Spatial Autocorrelation) Help

Tests whether a variable is spatially clustered using k-nearest-neighbour weights and a permutation p-value.

Needs: Latitude, longitude and a numeric value column.

Configure

Machine Learning

Random Forest Help

Multi-split Random Forest regression with feature importance, optional univariate screening, permutation tests and SHAP values.

Needs: One dataset with a numeric target column.

Configure
Model Comparison Help

Benchmark multiple regressors (linear, trees, boosting, SVR, neural net) under the same cross-validation splits.

Needs: One dataset with a numeric target column.

Configure
Full Pipeline Help

End-to-end run: preprocessing, feature engineering and selection, model training, evaluation and interpretation.

Needs: One dataset with a numeric target column.

Configure

Multi-Omics Integration

PCA (Multi-Omics) Help

Per-block principal component analysis with an integrated overview.

Needs: Multiple omics blocks (wide table or separate files).

Configure
SNF Help

Similarity Network Fusion — integrates blocks into one sample-similarity network.

Needs: Multiple omics blocks. Requires the snfpy package.

Configure
MOFA+ Help

Multi-Omics Factor Analysis — latent factors driving variation across views.

Needs: Multiple omics blocks. Requires the mofapy2 package.

Configure
MOFA2 Help

Unsupervised factor discovery across multiple omics views (full pipeline).

Needs: Multiple omics blocks. Requires the mofapy2 package.

Configure
DIABLO Help

Supervised multi-block discrimination of outcome classes (mixOmics).

Needs: Multiple omics blocks and a categorical outcome. Requires R + mixOmics.

Configure
CCA Help

Canonical Correlation Analysis — maximally correlated components across blocks.

Needs: Two or more omics blocks.

Configure
NMF Help

Non-negative Matrix Factorisation — parts-based decomposition per block.

Needs: Omics blocks with non-negative values.

Configure

Genomics & Pathway

GWAS Help

Genome-wide association study: genotype encoding, QC, LD pruning, association testing and Manhattan plot.

Needs: A genotype dataset (configured in the dataset editor).

Configure
Pathway Analysis Help

Map features to pathways and run enrichment analysis (Reactome, KEGG, GO, GSEA).

Needs: A dataset with mappable feature identifiers.

Configure