Feature Selector
Data Preparation
Import
Clean
Filter
Missing Values
Unsupervised Models
PCA & MDS
Supervised Models
GLM-Net
sPLS-DA
Multi-Block sPLS-DA
Random Forest
Explore Data
Univariate Analysis
GLM
Odds Ratios
Export Results
Import Data
Choose excel file
Browse...
Submit
Overview of the source data
Clean Data
Consider numeric-looking variable as factor if:
Maximal number (included) of levels:
Maximal number (included) of digits:
Submit
Automatic changes to the data
Read it attentively and modify heuristics to obtain the data that makes sense
Save
Filter Data
No response (for unsupervised only)
Filter rows (observations)
for response variables
Remove Filter
Add Filter
for predictors
Apply same filters
Filter columns (variables)
Overview of the selected data
Visualize NA's
Save
Save
png
jpg
svg
pdf
Missing Values
Stringent
Column (variable) stringent first
Maximal tolerated % of NA for the first stringent:
Maximal tolerated % of NA for the second stringent:
Imputation
Select imputation method:
k Nearest Neighbours
Central Tendency
Replace NA with 0
Submit
Stringent
Worst variables after stringent
Imputation effect on correlations
Overview of Imputed Data
Save
png
jpg
svg
pdf
Save
png
jpg
svg
pdf
Save
png
jpg
svg
pdf
Save imputation histograms (it may take a lot of time to generate)
Save
PCA & MDS
Label by:
variable
cluster (on first 10 dims max)
Only numeric
All to numeric (otherwise dummies for factors)
Scale only numeric (including factors converted to numeric)
Scale all (dummies included)
Select the model:
PCA
Metric MDS (PCoA)
Non-Metric MDS
Submit
Scree Plot
Scatter Plot
Density Plots
Save
png
jpg
svg
pdf
Save
Ellipses
Biplot
Density
Interactive
png
jpg
svg
pdf
Save (might be slow to generate gif!)
GLM-Net
Impose alpha
Standardize dummies
Choose the number of bootstraps:
Choose the oversampling:
Submit
Choose the threshold bootstrap frequency for the plot:
Bootsrapped plot
Bootsrapped FS table
FS on the original data
Save
png
jpg
svg
pdf
Save
Save model
Save table
sPLS-DA
Submit
Choose parameters for graphics
Feature stability threshold:
PCA
Regular PLS-DA
Optimal Parameters
sPLS-DA
Feature Stability
Save
Ellipses
Biplot
Density
Interactive
png
jpg
svg
pdf
Save model
Save plot
Ellipses
Biplot
Density
Interactive
png
jpg
svg
pdf
Save
Save model
Save plot
Ellipses
Biplot
Density
Interactive
png
jpg
svg
pdf
Save
png
jpg
svg
pdf
Multi-Block sPLS-DA
Submit
Choose cut-offs for graphics
PLS correlation circle:
DIABLO:
Feature stability threshold:
Circos:
Pairwise PLS
Optimal Parameters
DIABLO
Feature Stability
Block Interactions
Save
Interactive
png
jpg
svg
pdf
Save
Save
Ellipses
Biplot
Interactive
png
jpg
svg
pdf
Save
png
jpg
svg
pdf
Save
Random Forest
Use RF Imputation
Random Forest
Number of Trees:
Single Tree
Submit
Optimization
FS
MDS
Single Tree
Save
png
jpg
svg
pdf
Save model
Save plot
png
jpg
svg
pdf
Save
Interactive
png
jpg
svg
pdf
Save
png
jpg
svg
pdf
Univariate Analysis
Volcano plot parameters:
Submit
Download Current Plot
Max label overlap for volcano plot:
Plot
Volcano Plot
Save
Interactive
png
jpg
svg
pdf
Save
Interactive
Adjust p-values
png
jpg
svg
pdf
Save
Add Confidence Intervals
Interactive
png
jpg
svg
pdf
GLM
Here you can conrol for potential confounders!
Submit
GLM
Save model
Save message
Save
Add Confidence Intervals
Interactive
png
jpg
svg
pdf
Odds Ratios
Submit
Choose the significance level:
Forest Plot
OR by variables
Save
png
jpg
svg
pdf
Save
Export Results
Download outputs:
select all
deselect all
Apply export selection
Download Selection
Review Selection
Update export checkboxes
Data Preparation
Cleaning changes
Filtered data
Missing values after filters
Stringent plots
Missing values after stringent
Correlation heatmaps
Imputation histograms (it may take a lot of time to generate)
Imputed data (with response)
PCA & MDS
Scree plot
PCA / MDS plot
Density plots (might be slow to generate gif!)
GLM-Net
Bootstrap barplot
Feature counts
GLM-Net model
Table with selected features
sPLS-DA
PCA plot
PLS-DA model
PLS-DA plot
sPLS-DA tuning
sPLS-DA model
sPLS-DA plot
sPLS-DA feature stability
MB sPLS-DA (DIABLO)
Pairwise PLS plots
DIABLO tuning
DIABLO
DIABLO feature stability
Block interactions
Random Forest
RF tuning
RFmodel
RF feature importance
RF MDS
Single decision tree
Explore Data
Univariate plot
Volcano plot
ROC (for univariate analysis)
GLM (model)
GLM message
ROC (for GLM)
Forest plot
Odds ratios table