FAIRMol
Chemoinformatics Analysis
Export HTML Compounds Samples ← Compounds
Analysis views
Computing…
Projection

Projection workspace

Projection controls are now kept directly next to the graph: method settings on the left, the graph in the middle, and visualization controls on the right.
20% · 60% · 20% layout Submit sits under method settings
Choose the projection method and any method-specific parameters in the left column of the Projection view, then submit there to recompute the map.
Colouring, sizing, and selection actions now stay on the right side of the graph so display tweaks are separated from projection computation.
Open the Projection tab below to work in the new three-column layout with a less stretched central graph.
Histograms

Distribution overview

Histogram controls live inside the histogram cards below. Use this view when you want a quick sense of ranges, skew, outliers and multimodality across descriptors.
Per-property controls inside cards Click histogram → table below
Start here for a fast read on property spread before switching to Projection or Properties for deeper comparison.
Use the statistic chips and the histogram table together: the plot shows shape, while the list gives exact counts and navigation.
No global settings are required here. Choose the property to inspect directly inside the histogram view.
Scaffolds

Scaffold controls

Tune scaffold definition and how many top scaffold families are emphasised in the chart and table below.
Affects scaffold charts + cluster tree context
Recommended default: 12.
Broader scaffold variants merge more analogues; tighter variants preserve finer chemotype distinctions.
Functional groups

Motif-focused browsing

This module is primarily interaction-driven: use the bar chart and table below to jump into compounds sharing the same motif.
Top groups plotted automatically Chart + table stay in sync
Useful for spotting recurring medicinal chemistry motifs, privileged substructures, or overrepresented liabilities.
Use the chart for the dominant groups and the table when you need the long tail of less frequent motifs.
No global settings are needed here. Click a group to populate the shared compound selection panel in the module below.
Properties

Descriptor scatter guidance

Axis, colour and size controls are part of the scatter panel below, so this header stays focused on workflow guidance.
X/Y/colour/size inside scatter card Lasso selection enabled
Great for finding outliers, checking correlations, and comparing scaffold families in descriptor space.
Try MW vs LogP first, then colour by TPSA or scaffold family and use lasso to inspect the interesting corner of the map.
The property panel updates live inside the module; you do not need a page refresh for axis or colouring changes there.
Clusters

Cluster controls

Choose the clustering strategy and tune only the parameters for the active method. The tree and detail table below update from these settings.
Only active method shows parameters
Cluster tree labels use the current scaffold settings from the Scaffolds module.
Recommended default: 0.60.
Higher values create tighter similarity clusters.
Recommended default: 0.90.
Larger epsilon merges nearby points into broader clusters.
Recommended default: 8.
Use K-Means when you want a fixed number of groups for comparison.
Selections

Selection workbench

All saved subsets, temporary selections, set algebra and side-by-side comparison live in this dedicated tab, so the other analysis views stay cleaner.
Set A / B / C Temp subset workflow Compare selections
Open this tab after selecting compounds anywhere in the dashboard. Save the subset, compare it with another set, then rerun the analysis only on that focused library.
Histogram → capture current selection → save to Set A; Projection → capture chemotype cluster → save to Set B; then compare A vs B or intersect them.
Selection comparison currently uses scaffold, functional-group and descriptor summaries. It is intentionally lighter-weight than a full atom-level MCS workflow.
Loading analysis view…