Documentation
Field-by-field data reference
Definitions for every column across the three datasets behind KinaDB — the same reference used to build the tables, filters, and structure pages. See also the downloads page for the raw CSVs.
1. Browse table — Browse.csv
| Column | Meaning | UI treatment |
|---|---|---|
| pdb | 4-character PDB ID of the crystal / cryo-EM structure. | Monospace, links to detail page |
| chain | Chain identifier within that PDB entry (a single PDB can have multiple chains, each a separate row). | Monospace, small |
| Protein_name | Full descriptive protein name as deposited. | Plain text |
| kinase | Standardized kinase name / gene symbol (e.g. EGFR, BRAF, CDK2). | Filterable/searchable, links to filtered Browse |
| group | Kinase group per the standard kinome classification (AGC, CAMK, CK1, CMGC, STE, TKL, TYR, NEK, RGC, Other). | Dropdown filter / colored tag |
| species | Source organism (e.g. Human, Mouse). | Dropdown filter |
| ligand | Orthosteric ligand bound in the ATP pocket, if any. | Searchable, links to filtered Browse |
| allosteric_ligand | Ligand bound at a non-ATP-pocket (allosteric) site, if any. | Searchable |
| Predicted_Activity | Model-predicted activation state of the kinase in this structure (Active / Inactive). | Colored badge, dropdown filter |
2. Structure detail — Browse_each_pdb_structure.csv
Grouped exactly as the structure detail page presents them.
Header
| Column | Meaning |
|---|---|
| PDB ID | Structure identifier. |
| Chain | Chain within the structure. |
| Protein Name | Full protein name. |
| Resolution (Å) | Crystallographic resolution in ångströms — lower is higher quality/detail. |
| R-Value | Crystallographic R-factor — how well the atomic model fits the diffraction data; lower is a better fit. |
Kinase info
| Column | Meaning |
|---|---|
| Kinase | Standardized kinase name. |
| Kinase Group | Kinome group classification. |
| Species | Source organism. |
| UniProt ID | Cross-reference to the UniProt sequence/function database — rendered as an outbound link. |
Sequence info
| Column | Meaning |
|---|---|
| PDB Sequence | The amino-acid sequence actually observed/resolved in this structure. |
| UniProt Reference Sequence | The canonical full-length sequence for this protein from UniProt. |
| Aligned Reference Sequence | The UniProt reference aligned to the PDB sequence (same coordinate frame). |
| Projected Alignment | The alignment mapping/annotation projected onto the structure (positional correspondence). |
| Activation Segment | The activation-loop sequence stretch — a key regulatory region whose conformation determines the active/inactive state. |
| Missing Residues | Residues present in the reference but not resolved in this structure (common in flexible loops). Shaded against the reference in the sequence block. |
Structure info
| Column | Meaning |
|---|---|
| Binding Pocket | Residues/description of the ATP-binding pocket in this structure. |
| Alternate Conformation | Whether alternate side-chain/backbone conformations are modeled (common at high resolution). |
| Missing Atoms | Specific atoms (not full residues) unresolved in the electron density. |
| Salt Bridge (KinCore) | Presence/state of the regulatory Lys–Glu salt bridge per the KinCore method — a standard marker of activation state. |
| Consensus DFG Conformation | DFG-in vs DFG-out state of the DFG motif. DFG-in is generally active; DFG-out is inactive/allosteric-inhibited. |
| Consensus αC-Helix Conformation | αC-helix-in vs -out — another core marker of activation, independent of but related to DFG state. |
Ligand info
| Column | Meaning |
|---|---|
| Ligand | Orthosteric ligand name. |
| Allosteric Ligand | Allosteric-site ligand name. |
| Ligand Type | Category of the bound ligand (inhibitor type, cofactor, etc.), aligned with Inhibitor_Type in the Drugs dataset where possible. |
Activity prediction
| Column | Meaning |
|---|---|
| Predicted Activity | Model's classification of this structure's activation state. |
| Activity Probability | Model's confidence in that classification (0–1), shown as a probability bar next to the badge. |
3. Drugs table — kinase_inhibitor_drugs.csv
| Column | Meaning | UI treatment |
|---|---|---|
| Drug_Name | Primary name of the drug/compound. | Searchable, primary column |
| Synonyms | Alternate names/codes the drug is known by. | Searchable, secondary text |
| Target_Kinase_Gene_Symbol(s) | Kinase(s) this drug targets, by gene symbol. | Where a value matches a kinase in Browse, links to that kinase's filtered Browse results |
| Kinase_Group_KLIFS_style | Kinase group of the target, using KLIFS-style naming (broadly matches the Browse group field). | Dropdown filter |
| Approval_Status | Regulatory status (FDA-Approved, Investigational, Withdrawn, Clinical-trial phase, …). | Colored badge, dropdown filter |
| Indication | Disease(s)/condition(s) the drug is approved or investigated for. | Plain text, searchable |
| Inhibitor_Type | Mechanistic class of inhibitor (Type I, Type II, allosteric, covalent, …) — maps to the DFG/αC state it targets. | Dropdown filter |
Cross-dataset relationships
pdb + chain is the primary key joining Browse to the structure-detail records (one row per structure = one full detail page).
Browse.kinase and Drugs.Target_Kinase_Gene_Symbol(s) are matched on gene symbol to power the Drugs ↔ Browse cross-links.
Browse.group and Drugs.Kinase_Group_KLIFS_style represent the same underlying kinome classification (label sets may differ slightly).