Peptide 3D Structure from HELM

Peptide 3D Structure from HELM takes a peptide written in HELM notation and builds a three dimensional structure for it. A peptide is a short chain of amino acids, the same building blocks proteins are made of. HELM is a text format for spelling out exactly which amino acids are in the chain, in what order, and how they are joined together, including unusual joins like rings.

A HELM2 string looks like PEPTIDE1{A.G.F.K.L}$$$$V2.0, where the letters between the braces are the amino acids in order. HELM can describe more than a simple straight chain: it can spell out head to tail cyclization, where the two ends of the peptide are joined into a ring, disulfide bridges and lactam bridges, which are extra cross links inside the chain, and non canonical amino acids, meaning building blocks beyond the twenty standard ones.

The tool reads that description and produces a real 3D structure you can download and use, with one or more conformers ranked by energy. A conformer is one specific 3D arrangement of the chain, and the lower its energy the more stable and favorable that shape is.

When to use it

Use it when you have a peptide described in HELM and you need an actual 3D structure for it, for example to visualize it, dock it, or feed it into another calculation that needs real coordinates. It is especially useful for cyclic peptides and for peptides that contain non standard amino acids, which plain sequence tools often cannot build.

Inputs

Input	Required	What it is
`helm`	yes	HELM2 notation string describing the peptide, for example `PEPTIDE1{A.G.F.K.L}$$$$V2.0`.
`n_conformers`	no, default `1`, min `1`, max `100`	How many conformers to generate and rank by energy.
`conformer_method`	no, default `etkdg`, one of `etkdg` or `xtb`	How the 3D shape is built. `etkdg` is fast (about 3 seconds). `xtb` runs a GFN2-xTB quantum geometry optimization, which is slower (about 30 seconds) but gives a more refined shape.
`force_field`	no, default `MMFF94s`, one of `MMFF94s` or `UFF`	The force field used to optimize the geometry with the `etkdg` method. Ignored when the method is `xtb`.

How to run it

Submit your peptide from Azulene Studio, the Python SDK, or the CLI. New here? The Get started page walks through installing, logging in, and running a ready made example first.

In Azulene Studio

Open Peptide 3D Structure from HELM from the tools list, then on the Inputs and Parameters step enter your HELM string, set how many conformers you want, and pick the method and force field if you want to change the defaults, then Review and Submit.

From the Python SDK

from opal import jobs

result = jobs.submit(
    job_type="peptide_structure",
    input_data={
        "helm": "PEPTIDE1{A.G.F.K.L}$$$$V2.0",
        "n_conformers": 1,
        "conformer_method": "etkdg",
    },
)

From the CLI

Pass the inputs as a JSON string.

opal jobs submit --job-type peptide_structure \
  --input-data '{"helm": "PEPTIDE1{A.G.F.K.L}$$$$V2.0", "n_conformers": 1, "conformer_method": "etkdg"}'

Reading the result

The result summarizes the assembled peptide and gives you the structure to download:

molecular_weight: the monoisotopic mass of the assembled peptide, in daltons (Da).
n_residues: how many amino acid residues are in the peptide.
validation.ring_closure_quality: for a cyclic peptide, a score for how cleanly the ring closes, where 1.0 is ideal geometry. Linear peptides have no ring to close, so this is blank for them.

The downloadable 3D structure file (PDB or SDF) is available on the Files tab of the result in Azulene Studio.

Notes

The etkdg method is fast and good for a quick structure. Choose xtb when you want a more carefully optimized geometry and are willing to wait longer. Asking for more conformers samples more possible shapes of the chain and ranks them by energy, with lower energy meaning a more stable shape.