--- title: "How to Create a Peptide Dataset" slug: "how-to-create-peptide-dataset" status: "update" updated: 2026-10-14T09:21:42Z published: 2026-10-14T09:21:42Z canonical: "docs.discngine.com/how-to-create-peptide-dataset" --- > ## Documentation Index > Fetch the complete documentation index at: https://docs.discngine.com/llms.txt > Use this file to discover all available pages before exploring further. # How to Create a Peptide Dataset This guide covers importing a peptide dataset from a HELM file (CSV/Excel) or an SDF, supplying a monomer library, and configuring peptide-specific options. Peptide plugin required The **Peptide** dataset type is only available when the peptide plugin is enabled for your organization. If you do not see it, contact your administrator. ## Step 1: Open the New Dataset wizard From the **Home** page, click the **Create Dataset** card (or **New → Create Dataset** in the left navigation). The **New Dataset** wizard opens with three steps: **Dataset Information**, **Properties Configuration**, and **Import Status**. ## Step 2: Choose the project and the dataset file Choose the target **Project**, then drop your file on **Dataset File** (or click it to browse). Ideation reads the file, sets the **Dataset Type** for you and tells you when it switched it, with *"Detected a Peptide dataset. You can change the type above."* ![The New Dataset dialog after dropping the peptide CSV template: Project, then the Dataset Type switched to Peptide, the dropped file, and the peptide template download links](https://cdn.document360.io/dde9293e-c107-4b3e-9774-9c9c296ed970/Images/Documentation/create-dataset-peptide-type.png?sv=2026-02-06&spr=https&st=2026-10-27T12%3A05%3A20Z&se=2026-10-27T13%3A05%3A20Z&sr=c&sp=r&sig=bY36SSZkGipthOOgjvB18dssBjxDTfTcNJKQDSd0VmY%3D) | File you drop | Dataset Type selected | | --- | --- | | CSV with a **HELM** column | **Peptide** | | CSV without a HELM column | **Small Molecule** | | Excel (`.xlsx`) | **Peptide** | | SDF whose first molecule carries monomer templates | **Peptide** | | Any other SDF | **Small Molecule** | | Compressed file (`.csv.gz`, `.sdf.gz`, `.sd.gz`) | Not read: the selected type is kept | Check that **Dataset Type** reads **Peptide** before you continue, and select it yourself if the guess is wrong or your file is compressed. If the type does not accept the file's format (an Excel file with **Small Molecule**, for example), **Next** stops with *"This file format is not supported for Small Molecule datasets."* Accepted file formats for peptides: | Format | Notes | | --- | --- | | CSV / Excel (`.csv`, `.csv.gz`, `.xlsx`) | Tabular files with an identifier column and a **HELM** column; every other column becomes a property. HELM references monomers by **symbol** only, so a **monomer library** is needed for any monomer not already in the Biotoolkit Monomer Service. | | SDF (`.sdf`, `.sd`, `.sdf.gz`, `.sd.gz`) | A V3000 SDF that carries the peptides **and** a monomer template for each monomer. The embedded templates are used to register any missing monomers on the Biotoolkit Monomer Service, so **no separate monomer library is needed**. | Start from a template With the **Dataset Type** set to **Peptide**, use the **Download peptide dataset templates** links (CSV / Excel) under the file picker to start from a correctly formatted file. The template's columns are `ID`, `HELM` and three example property columns, so a minimal row looks like `PEP001,PEPTIDE1{A.G.K.L}$$$$`. For branches, bridges, chemical objects and the HELM features Ideation does not support, see [HELM Rules for Peptide Import](/public/docs/reference-helm-rules). ## Step 3: Supply a monomer library (CSV/Excel imports only) A **Monomer Library File** maps each monomer **symbol** to its chemical structure so Ideation can register monomers that are not yet in the **Biotoolkit Monomer Service**. It is needed only for **CSV/Excel (HELM) imports** — HELM references monomers by symbol without carrying their structures — and the picker appears only for those files. ![The New Dataset dialog showing the selected CSV peptide file and the optional Monomer Library File picker with CSV, Excel and SDF template links](https://cdn.document360.io/dde9293e-c107-4b3e-9774-9c9c296ed970/Images/Documentation/create-dataset-monomer-library.png?sv=2026-02-06&spr=https&st=2026-10-27T12%3A05%3A20Z&se=2026-10-27T13%3A05%3A20Z&sr=c&sp=r&sig=bY36SSZkGipthOOgjvB18dssBjxDTfTcNJKQDSd0VmY%3D) The notice above the picker sums up how the upload relates to the Biotoolkit Monomer Service: - Monomers live in the **Biotoolkit Monomer Service**. If every monomer of your dataset is already registered there, you do not need a library file at all. - **Only the monomers your dataset actually uses are registered** from the file. A shared library of thousands of monomers is fine: the rows your HELM sequences never reference are skipped, and skipping them is not reported as a problem. - You can also upload the same file **directly in the Biotoolkit Monomer Service app** — the notice links to it — instead of attaching it here. It can be a CSV, Excel, or SDF file, and it uses **the same field names as the Biotoolkit Monomer Service import template** — so a library file written for the Biotoolkit Monomer Service imports into Ideation unchanged, and vice versa. | Field | Required | What it holds | | --- | --- | --- | | `symbol` | yes | The symbol the HELM sequences use. | | `smiles` | one of | The structure as SMILES, with numbered attachment points — `C[C@H](N[*:1])C([*:2])=O`. | | `ctab` | one of | The structure as a connection table. For an SDF, the molfile itself is the structure and no `ctab` field is needed. | | `name` | no | Falls back to the symbol. | | `polymer_type` | no | `PEPTIDE` or `CHEM`. Inferred from how the dataset's HELM uses the monomer when absent. | | `monomer_type` | no | `Backbone` or `Undefined`. | | `natural_analog` | no | The natural amino acid the monomer stands in for. | | `rgroups` | no | The attachment points and their cap groups, as `R1:H;R2:OH`. Derived from the structure when absent. | | `R1`, `R2`, … | no | The same declaration spread over one column per attachment point, each holding its cap group (`H` or `OH`) — the way some registration systems export it. Ignored when `rgroups` is also stated. | | `author` | no | Accepted for compatibility with the Biotoolkit Monomer Service template; not stored. | Each row needs a `symbol` and a structure — **either** `smiles` **or** `ctab`. Columns Ideation does not know are ignored, so extra columns are never a reason for a file to fail. Field names are matched on their letters only: `natural_analog`, `Natural Analog` and `NaturalAnalog` all name the same field, whatever the casing or the separator. Exports from a registration system that writes CamelCase headers therefore import as they are. One bad row does not fail the file A row Ideation cannot read — an unknown polymer type, a row with neither `smiles` nor `ctab`, a SMILES whose attachment point carries no R-number or that marks its stereo as relative or racemic (`|&1:…|`, `|o1:…|`) when no `ctab` is there to fall back on — is **skipped and reported**, while every other row imports. Skipped rows appear in the dataset's import issues dialog, under **Monomers not registered**, as *Monomer "X" (library row N)* with the reason, but only when a peptide of the dataset actually uses that monomer; the peptides built from it are listed under **Peptides imported with missing data** as missing that monomer. A row that reads fine but that the Monomer Service refuses — invalid chemistry, a structure already registered under another symbol — is reported the same way, with the service's reason. Two smaller problems cost less than the row: an unreadable `rgroups` or `R1`… declaration is ignored and the cap groups are derived from the structure instead, and an unusable `smiles` is dropped when the row also carries a `ctab`, which then provides the structure. Both are recorded in the server log only. When you state a field, Ideation trusts it `polymer_type`, `monomer_type` and `rgroups` are the fields Ideation can work out on its own — but when your file states them, your values win over what Ideation would have derived from the structure. That is what lets you register a monomer whose structure carries no usable attachment-point labels; it also means an incorrect value registers an incorrect monomer, so state them only when you mean to. Older library files still import Files written against the previous Ideation template keep working: the old spaced `Natural Analog` header is still accepted (see the matching rule above), `CTAB` still matches, and the now-obsolete `Short Name` column is simply ignored. No monomer library needed for SDF For **SDF** dataset imports you do not provide a monomer library. A V3000 SDF already embeds a monomer template for every monomer, and those templates are used to register any missing monomers on the Biotoolkit Monomer Service automatically — which is why the picker does not appear for SDF files. ## Step 4: Configure peptide-specific options Click **Next**. On the **Properties Configuration** step, a **Peptide Specific Configuration** section appears: ![The Properties Configuration step showing the Peptide Specific Configuration section with the HELM Column and Reference Peptide selectors](https://cdn.document360.io/dde9293e-c107-4b3e-9774-9c9c296ed970/Images/Documentation/create-dataset-peptide-config.png?sv=2026-02-06&spr=https&st=2026-10-27T12%3A05%3A20Z&se=2026-10-27T13%3A05%3A20Z&sr=c&sp=r&sig=bY36SSZkGipthOOgjvB18dssBjxDTfTcNJKQDSd0VmY%3D) - **HELM Column** — the column that holds the HELM sequence notation (tabular imports only; auto-detected when named `HELM`). - **Reference Peptide** — the peptide used by default as the basis for SAR slides created from this dataset. You can change it later per report. Select and configure at least one main property, then submit. ## Step 5: Watch the import The final **Import Status** step shows the peptide import timeline: **Initializing → Registering Monomers → Parsing Peptide Sequences → Ready** ![The Import Status step showing the four-stage peptide import timeline ending in Ready, with a banner reporting the structures imported with missing data](https://cdn.document360.io/dde9293e-c107-4b3e-9774-9c9c296ed970/Images/Documentation/import-status-timeline-peptide.png?sv=2026-02-06&spr=https&st=2026-10-27T12%3A05%3A20Z&se=2026-10-27T13%3A05%3A20Z&sr=c&sp=r&sig=bY36SSZkGipthOOgjvB18dssBjxDTfTcNJKQDSd0VmY%3D) - **Registering Monomers** runs when a monomer library was supplied, or when the dataset is a **V3000 SDF** (whose embedded monomer templates are registered on the Biotoolkit Monomer Service). When the dataset reaches **Ready**, it opens on the dataset page. Its alignments are computed next, in the background: a **Center-star alignment** and a **PepSeA alignment** appear on the dataset's **Alignments** tab and complete on their own, so you can build a SAR Slide without providing an alignment file. Existing monomers are never modified Registration only adds monomers that are **new** to the Biotoolkit Monomer Service. Monomers whose symbol already exists there are used as-is and are **never overwritten or updated** — so a library or SDF cannot change a monomer that other datasets already rely on. The service also runs a **duplication check on structure**: if the same monomer is already registered under a different symbol, it is matched to that existing entry (as an alias) rather than registered again. Automatic alignment is best-effort An alignment that cannot run ends as a **Failed** row on the **Alignments** tab, and the import still succeeds. Center-star does not run on datasets with no HELM sequences or above 5,000 peptides; PepSeA does not run on datasets with too many distinct non-natural monomers. See [How to Align Peptides](/public/docs/how-to-align-peptides). Some peptides may import with missing data If some peptides reference monomers that could not be resolved, they are imported with their HELM sequence preserved but without full structure or properties, and the dataset is flagged with an amber import issues warning that lists them under **Peptides imported with missing data**. See [How to Resolve Missing Monomers](/public/docs/how-to-resolve-missing-monomers). ## Step 6: Read the dataset table The dataset page lists one row per peptide. Besides your own properties, a peptide dataset carries these common columns: | Column | Shown by default | What it holds | | --- | --- | --- | | **ID** | yes | The primary identifier. | | **Backbone Sequence** | yes | The peptide drawn as its sequence of monomers. | | **Num Monomers** | yes | How many monomers the HELM declares. | | **Num Atoms** | yes | Atom count of the converted structure. | | **Molecular Weight** | yes | Computed from the converted structure. | | **HELM** | yes | The peptide's HELM notation. When the conversion succeeds and no monomer had to be inlined, its chains are renumbered in the order your file writes them, so a chain number can differ from your file — see [HELM Rules for Peptide Import](/public/docs/reference-helm-rules). Otherwise, the HELM is stored as written. | | **Canonical SMILES** | no | Canonical SMILES of the whole peptide, computed from the converted structure. | | **Inchi Key** | no | InChIKey of the whole peptide, computed from the converted structure. | | **Missing Monomers** | when needed | The unresolved symbols of a peptide imported without its structure; the column exists only while something is unresolved. | Tick **Canonical SMILES** and **Inchi Key** in **Table settings › Columns** to show them; they start unticked because they are long identifiers you copy rather than read. Both are empty for a peptide whose structure could not be converted, and for peptides imported before this release until you run **Sync monomers** (see [How to Resolve Missing Monomers](/public/docs/how-to-resolve-missing-monomers)). ![The Table settings drawer of a peptide dataset, Columns tab, with Canonical SMILES and Inchi Key listed but unticked](https://cdn.document360.io/dde9293e-c107-4b3e-9774-9c9c296ed970/Images/Documentation/peptide-table-settings-columns.png?sv=2026-02-06&spr=https&st=2026-10-27T12%3A05%3A20Z&se=2026-10-27T13%3A05%3A20Z&sr=c&sp=r&sig=bY36SSZkGipthOOgjvB18dssBjxDTfTcNJKQDSd0VmY%3D) The row menu (the list icon in the **Actions** column of a row) offers **Create SAR Slide** with that peptide as reference, and copies to the clipboard: **Copy HELM**, **Copy CTAB v2000**, **Copy CTAB v3000**, **Copy Canonical SMILES** and **Copy InChIKey**. A copy is disabled when the peptide has no such value. ## Next steps - [HELM Rules for Peptide Import](/public/docs/reference-helm-rules) - [How to Resolve Missing Monomers](/public/docs/how-to-resolve-missing-monomers) - [How to Align Peptides](/public/docs/how-to-align-peptides) - [How to Create a Peptide SAR Report](/public/docs/how-to-create-peptide-sar-report)