UniProt to PDB SIFTS Residue Numbering and Chain Mapper
Resolve coordinate offsets between canonical UniProtKB sequence positions, mature cleaved-protein numbering, and PDB author chain labels using human EGFR (P00533 / PDB 3IKA, a kinase-domain structure). In structural biology, subtracting a fixed integer offset is never a safe substitute for residue-level SIFTS alignment because expression tags, internal loop deletions, and insertion codes break simple arithmetic.
Demonstrates explicit UniProt-to-PDB coordinate reconciliation on EGFR P00533. PDB 3IKA is a kinase-domain construct (SIFTS chains A and B, UniProt 694–1022) carrying T790M, so UniProt residue 790 maps to a resolved methionine there, unlike the mature-peptide numbering (−24 signal peptide) that would call it 766.
Why arithmetic offsets fail across UniProt, mature peptide, and PDB coordinates
Literature and database records routinely refer to the same physical amino acid by two or three incompatible numbers. In human epidermal growth factor receptor (UniProt P00533), the canonical database sequence includes the 24-residue N-terminal signal peptide, placing the kinase gatekeeper threonine at T790 and the sensitizing leucine at L858. Early structural and biochemical studies numbered residues from the cleaved mature N-terminus (shifting every coordinate by −24 to T766 and L834), while individual PDB depositions of isolated kinase constructs may start numbering at 1 for the N-terminal His-tag or contain engineered internal loop deletions.
Because crystallography constructs frequently substitute solubility tags, delete flexible linkers, or apply family-wide numbering systems (such as Kabat/IMGT in antibodies or Ballesteros-Weinstein in GPCRs), applying a single global integer offset across a chain is error-prone. Only a residue-by-residue mapping table—such as the PDBe/UniProt SIFTS alignment—reliably connects each canonical UniProt coordinate to its corresponding PDB chain, author residue number, and insertion code.
Distinguish uncloned domains from crystallographically unobserved loops
When inspecting EGFR P00533 alongside PDB 3IKA in the protein workspace, notice that the experimental structure spans the intracellular kinase domain (UniProt 694–1022) while omitting the extracellular ligand-binding regions and transmembrane helix. Within the kinase domain itself, individual surface residues may also be unmodeled where local disorder prevented atomic coordinate refinement.
Before reporting a structural distance or highlighting a clinical variant on a 3D model, verify three facts in order: (1) that the canonical UniProt sequence has the stated wild-type amino acid at the named coordinate, (2) that the SIFTS mapping assigns that UniProt coordinate to a resolved residue in the chosen PDB chain, and (3) whether the PDB construct itself carries any engineered stabilizing mutations near the site.
Equations and variables
Precursor vs mature signal-peptide cleaved coordinate relation
pUniProt = pmature + Lsignal (for EGFR P00533: 790 = 766 + 24)
- pUniProt
- One-based residue index on the canonical full-length UniProt precursor sequence
- pmature
- Historical mature-peptide residue number after signal-peptide cleavage
- Lsignal
- Length of the cleaved N-terminal signal peptide (24 aa in human EGFR P00533)
Piecewise SIFTS residue mapping function
MapSIFTS : (UniProt_ID, pUniProt) → (PDB_ID, Chain_ID, auth_seq_id, ins_code) ∪ {Unobserved}
- auth_seq_id, ins_code
- Author-assigned PDB residue number and optional insertion letter (e.g. 100A)
- Unobserved
- Residues omitted from the expression construct or disordered in the electron density map
Scope and limitations
- A constant offset formula fails whenever a crystallized construct contains internal loop deletions, chimeric fusion partners (such as T4 lysozyme), or Kabat/Chothia insertion codes.
- A residue present in the crystallized construct SEQRES header may still lack 3D atomic coordinates (ATOM records) if local flexibility prevented electron-density resolution.
- Isoform-specific splicing changes UniProt residue indices downstream of the alternative exon boundary; always record the exact UniProt isoform suffix (e.g. P00533-1).
Sources and provenance
- UniProtKB human EGFR P00533CC BY 4.0
The interactive workspace displays the release of the retrieved record.
- Dana et al. 2019 SIFTS: updated Structure Integration with Function, Taxonomy and Sequences resourceCC BY 4.0; residue-level UniProt-to-PDB mapping architecture
- RCSB PDB experimental entry 3IKA: EGFR 696–1022 T790M with WZ4002CC0; wwPDB coordinate and SEQRES record
Technical questions
Why is the EGFR gatekeeper mutation written as T790M in modern clinical records but T766M in older biochemical papers?
Full-length human EGFR (UniProt P00533) is synthesized as a 1,210-amino-acid precursor whose first 24 residues form a cleavable N-terminal signal peptide. Counting from the initiator methionine places the gatekeeper threonine at position 790, whereas counting from the N-terminus of the mature 1,186-residue cell-surface receptor places the exact same amino acid at position 766.
What is the difference between label_seq_id and auth_seq_id in PDBx/mmCIF files?
In PDBx/mmCIF format, label_seq_id is a strict 1-to-N sequential index along the crystallized polymer entity (SEQRES), whereas auth_seq_id preserves the author-chosen residue number, which may follow canonical UniProt numbering, mature protein numbering, or a homologous family numbering scheme with lettered insertion codes.
Why can a residue appear in the PDB sequence alignment yet have no atom coordinates in the 3D viewer?
The SEQRES record lists every residue chemically present in the recombinant protein placed into the crystallization drop. If a surface loop or terminus is mobile in the crystal lattice, its electron density is smeared out and structural biologists omit those atoms from the coordinate table rather than guessing their positions.
Connected workflows
- EGFR sequence and structure inspection
Open the primary EGFR P00533 sequence-and-structure workbench.
- Six-frame translation and ORF inspection
Review how nucleotide reading frames define primary polypeptide coordinates.
- Protein A280 concentration calculation
Calculate protein extinction coefficients from the exact recombinant construct sequence.