Understanding Peptides in Ideation

This page explains the concepts behind the Peptide feature: how a peptide report is organized, how peptides are represented on the canvas, and what the SAR analysis tools (hotspots, synergies, clustering) tell a medicinal chemist. For step-by-step instructions, see the how-to guides linked at the end.

SAR Report vs SAR Slide

A SAR Report is the workspace that lives inside a project alongside your datasets. It is the object you share and manage access to.

A SAR Report contains one or more SAR Slides. A slide is a single view of the data — for a peptide report, each slide is one interactive viewer built around a reference peptide and an alignment. When you create a report you generate its first slide (the Generate SAR Slide button); you can add more slides later with Add SAR Slide — for example to compare a different reference peptide or a different alignment side by side.

Report vs Slide

"SAR Report" is the container; "SAR Slide" is one tab inside it. They are not the same thing.

How peptides are represented

HELM notation and monomers

Peptides are stored using HELM notation (Hierarchical Editing Language for Macromolecules), for example PEPTIDE1{H.S.Q.G.T.F...}$$$$. Each token in the sequence is a monomer — a natural amino acid (A, G, K, …) or a non-natural / modified residue (Aib, am, a PEG linker, and so on).

For Ideation to draw and compute properties for a peptide, it must know the chemical structure of every monomer symbol used in the sequence. That knowledge comes from two places:

Concept	What it is
Monomer library	A file (CSV, Excel, or SDF) you upload alongside a CSV/Excel (HELM) dataset, mapping custom monomer symbols to their chemical structure (CTAB/molblock) plus optional metadata. Not needed for SDF imports (see below).
Biotoolkit Monomer Service	A shared, organization-wide registry of monomers. Ideation resolves each symbol against it; symbols it cannot resolve are reported as missing. Registrations are permanent and shared across all datasets, so each non-natural monomer only needs to be registered once.

Where the structures come from depends on the dataset file:

CSV/Excel (HELM) files reference monomers by symbol only, so the structures for any non-standard monomers must come from an uploaded monomer library.
V3000 SDF files carry the peptides and a monomer template for each monomer, so no separate library is needed — the embedded templates supply the structures.

On import, Ideation resolves every symbol against the Biotoolkit Monomer Service. If a symbol is unknown but a structure is available (from an uploaded monomer library or an SDF's embedded templates), it is registered on the service automatically — only new monomers are added; symbols that already exist are used as-is and never overwritten. Symbols with no available structure are left missing — the peptide is preserved but its full structure and calculated properties are deferred until the monomers are registered and the dataset is synced. See How to Resolve Missing Monomers.

Backbone, branches, bridges and chemical objects

The viewer draws the reference peptide as a chain of monomer nodes and layers several structural features on top:

Element	Meaning
Backbone	The main linear chain of monomers, numbered by position.
Branch	A side chain that leaves the backbone at a given position and runs as its own numbered chain. Branches are drawn above the backbone with a red connector and a "Branch" label.
Bridge	A bond between two monomers (disulfide, lactam, thioether, …), drawn as a colored arc. A bridge can also run through a chemical linker, drawn as a node at the arc's middle.
Chemical object	A non-peptide chemical modification attached to the peptide (for example a PEG or biotin linker).
Cyclization	A ring layout for a run of monomers joined end-to-end by a bridge, or for a fully cyclic peptide.

Elements that appear in other peptides but not the reference (for example a bridge only some analogs have) are drawn dotted/faded and can be toggled on or off.

Alignment

Because peptides differ in length and composition, they are compared position-by-position through an alignment. Every SAR slide is built on an alignment that assigns each peptide's monomers to shared positions, inserting gaps where a peptide has nothing to contribute. An alignment belongs to the dataset, and a dataset can hold several: each slide references the one it was built on, so you can compare two framings of the same series side by side.

An alignment either comes from a file you upload (CSV for the backbone, Excel for backbone plus branches) or is computed by Ideation — which happens automatically for every dataset at import, and on demand whenever you ask for it.

The computation is a center-star multiple sequence alignment: every peptide is aligned against one shared reference — the dataset's longest backbone — and those pairwise results are stacked into a single frame. Monomer swaps are scored by how conservative they are rather than as a plain same/different test, so a modified residue lines up with the amino acid it derives from instead of counting as unrelated to everything.

For the full method — how the centre is chosen, how substitutions are scored, how gap costs are derived — see Understanding Peptide Alignment. For the task itself, see How to Align Peptides.

Understanding the SAR analysis

Reference peptide and variants

The reference peptide is the baseline the slide is built around. At each position (or bridge, region, or chemical object) the viewer collects the variants — the different monomers observed across all aligned peptides at that spot — and shows how many peptides carry each one, along with aggregated endpoint values. This is what a gallery displays when you open one.

Hotspots and synergies

Hotspot analysis scores how much each position (and optionally each bridge, chemical side chain, or region) impacts a chosen endpoint. Each element gets an impact bar (0–100%), a significance badge of Hot, Warm or Cold, and gallery deltas showing how each variant shifts the endpoint. Hotspots answer: which positions matter most for this property?

A single position rarely tells the whole story, so multi-position synergy analysis looks at combinations of positions — pairs, triples, or contiguous regions — and detects when changing them together departs from what their individual effects predict, either amplifying (synergistic) or cancelling (antagonistic). Each detected synergy appears as a region on the viewer and in a sortable list.

Both scores are relative, so some element always reads 100%; and both are built on the same per-position effects, which is why they are explained together in Understanding Hotspots and Synergies.

Clustering

Clustering groups the peptides by overall structural similarity. You choose how much weight to give the backbone, branches, bridges, and chemical objects, and either fix the number of clusters or let Ideation auto-detect the best number. The Cluster Explorer then lets you inspect the groups — most importantly through the Summary table, which shows each cluster's consensus pattern, quality, size, and property distributions side by side. Clustering answers: which families of analogs exist in this series, and how do their properties differ?

The grouping uses the same graded monomer-substitution model as the alignment, and the scatter plot is computed differently from the grouping itself — see Understanding Peptide Clustering.

Documentation Index