Pre-generated antigen constructs and design heuristics

How to choose among pre-generated constructs

Start with the protein region your experiment requires, then compare the evidence supporting its boundaries. The categories describe how boundaries were chosen; they do not rank expected expression success.

1. 1. Define your regionDecide whether you need the complete soluble region, a specific domain, or a membrane protein.
2. 2. Check experimental precedentLook for a relevant PDB structure and review the construct described in the associated study.
3. 3. Compare candidatesIf the best boundaries are unclear, compare a larger region with a smaller candidate.
4. 4. Inspect and exportUse the builder to inspect the boundaries and warnings before exporting your sequence.

A relevant PDB structure can guide your first candidate. Check the original study for fusion proteins, affinity tags, mutations, and required partners that may be absent from the mapped target sequence.

Before export

• Inspect both ends of the construct and retain any required domains or partners. Predicted boundaries may cut through helices or strands and do not establish that the isolated region will fold independently.
• Check whether truncation removed a cysteine's disulfide partner. Surface exposure alone does not justify mutating a cysteine, and burial does not establish that a substitution is safe.
• Review PTMs, processing sites, and relevant ortholog mappings. Use warnings to identify features that need closer inspection before deciding whether to keep or change the construct.

More detail: construct categories · using the builder · cysteines and processing.

After export

The exported sequence contains your selected target region and chosen edits. Choose a vector and host appropriate for your experiment, adding a secretion leader or tags where needed. Check whether native signal or processing sequences are already included.

OpenAntigens supports construct selection and design but does not prescribe expression or purification conditions. These practical guides can help plan the experimental work:

• General starting strategy: Structural Genomics Consortium et al., Protein production and purification (2008).
• E. coli expression: Aguilar Lucero et al., Starting a new recombinant protein production project in Escherichia coli (2021).
• Expi293F expression: Hederoth et al., Updated transient gene expression protocol in Expi293F cells using PEI (2025).
• ExpiCHO and HEK293E expression: Hacker et al., Small-scale cultivation and transfection of ExpiCHO and HEK293E cells in single-use orbitally shaken bioreactors (2024).
• Insect-cell expression: Jarvis, Recombinant protein expression in baculovirus-infected insect cells (2014).
• Tag selection: Kimple et al., Overview of affinity tags for protein purification (2013).
• Purification strategy: Wingfield, Overview of the purification of recombinant proteins (2015).

For tag choices, see Addgene’s protein-tag guide.

What the construct list contains

Each report lists proposed constructs with their sequences and the evidence used to choose the boundaries. The available classes depend on the annotations and structures found for that target.

Before ordering a sequence, check the target literature, required partners, and the assay you plan to run.

Strict and lenient construct algorithm

Calculated constructs use a compatible AlphaFold model and PAE matrix. Strict constructs use conservative pLDDT seeds followed by recursive PAE boundary scoring; lenient constructs use a lower pLDDT threshold to preserve larger structured regions.

Both tracks begin with intervals selected by pLDDT. The lenient track keeps these intervals; the strict track can split them further using PAE.

The steps below use the default configuration. A release built with different settings can produce different boundaries.

Step-by-step strict and lenient generation

1. Define the eligible design region. The pipeline first determines the region that can be used for soluble construct design. For secreted proteins this is usually the mature secreted chain; for single-pass proteins it is usually the extracellular region; for multipass proteins, large extracellular regions can be treated as mixed soluble/membrane cases, while membrane-expression constructs are handled separately. Strict and lenient soluble constructs stay inside this region.
2. Retrieve canonical AlphaFold confidence data when available. The canonical UniProt sequence is matched to the AlphaFold structure and PAE matrix. pLDDT is used as a per-residue local-confidence signal. PAE is used as a pairwise confidence signal that indicates whether two parts of the model have a confident relative orientation.
3. Generate lenient pLDDT regions. The lenient track marks residues with pLDDT at least 60, allows low-confidence gaps up to 12 residues, and discards resulting regions shorter than 25 residues. This track preserves larger structured units and can include more than one domain when the intervening linker is short or only moderately uncertain.
4. Generate strict pLDDT seed regions. The strict track marks residues with pLDDT at least 70, allows low-confidence gaps up to 8 residues, and discards seed regions shorter than 25 residues. These high-confidence intervals then feed PAE-based domain splitting.
5. Evaluate PAE split candidates inside each strict seed. For every possible cut, both resulting sides must remain at least 40 amino acids. The pipeline computes intrablock PAE within each side, interblock PAE between the two sides, and separation, defined as interblock PAE minus mean intrablock PAE.
6. Apply the PAE thresholds. A cut is eligible only when interblock PAE is at least 12 angstroms and separation is at least 4 angstroms. These thresholds select cuts where the model is less certain about placement between the blocks than within them.
7. Score eligible strict split boundaries. Candidate boundaries are ranked using split score = separation + 0.35 x max(0, boundary PAE - intrablock PAE) + linker bonus. The linker bonus increases when local pLDDT near the cut is low, because low-confidence linker residues support splitting adjacent structured blocks.
8. Recursively split strict regions. The highest-scoring eligible cut is applied, then the same PAE split search is repeated on the left and right child regions. Recursion stops when no valid cut remains, child regions would become too small, or the configured maximum recursion depth is reached.
9. Apply final construct filters. Calculated soluble constructs must stay inside the design region and must be at least 50 amino acids. Exact duplicate boundaries from different calculated or annotated sources are collapsed to keep reports concise; PDB-backed constructs are exempt from deduplication but still pass the scope and length filters.
10. Annotate each construct for review. Construct cards list sequence, boundaries, exports, included UniProt/InterPro annotations, PTMs, furin-like motifs, cysteine warnings, homolog-equivalent regions, and construct-specific conservation. Structural metrics, PAE split diagnostics, ligand-interaction annotations, and images appear when available.

Construct classes

Constructs are grouped by how their boundaries were chosen. The table follows the Construct Details tab order.

The default minimum construct length is 50 amino acids, including PDB-derived suggestions. Experimental PDB records also appear separately in the report.

AlphaFold-derived structural heuristics

Construct cards report the following metrics when the required structure data are available.

Interpret these diagnostics with domain annotations, PDB evidence, PTMs, ligand sites, cysteines, and the intended antibody-discovery strategy.

Domain and PDB evidence

What each construct card includes

Each card contains the primary human sequence and boundaries, with TSV/FASTA exports. Mapped homologs, annotations, and structural metrics appear when available.

Limitations

The generator proposes sequence boundaries; it does not predict expression yield, folding, or antibody binding. Review the target literature and test the selected constructs.