PROTAC design on the platform
Targeted protein degradation tools for PROTAC linker enumeration, ternary complex modeling, and degradation efficiency prediction.
Design bifunctional degraders that induce targeted protein degradation through the ubiquitin-proteasome pathway — enumerate linkers between a warhead and an E3 ligase ligand, model the three-body ternary complex for geometric feasibility, and predict degradation efficiency (DC50, Dmax, hook effect risk). Reach for these tools when your modality is targeted degradation rather than a conventional inhibitor. PROTAC design sits between an identified warhead and lead selection, feeding downstream into ADMET, retrosynthesis, and docking validation. See the Targets & Modalities overview for where this fits.
Backend parameters are narrower than the tool interfaces suggest
The Modal backend (PROTACParameters) only actually reads warhead_smiles, e3_binder_smiles, and num_linkers — all three workflows key off just those two SMILES strings. protac_linker_enumerate's linker_type and length_range are accepted by the tool but have no backend field to land in, so they're dropped before the job runs and don't affect which linkers come back (confirmed by an in-repo fix note, RQ-2308, that already caught and partially corrected this class of drop for this same tool). protac_ternary_model's target_pdb_uri/e3_pdb_uri/protac_smiles/num_poses and protac_degradation_score's protac_smiles/target_name/e3_ligase/cell_line also don't match any field the backend model defines — treat structure-based ternary modeling and named-E3/cell-line-context degradation scoring as not yet wired end to end, not as delivered capabilities.
PROTAC design overview
The PROTAC toolkit provides a computational pipeline for designing Proteolysis-Targeting Chimeras -- bifunctional molecules that induce targeted protein degradation via the ubiquitin-proteasome pathway. The three tools cover the full design cycle: linker chemistry exploration, ternary complex feasibility assessment, and degradation efficiency prediction.
Architecture
Tool inventory
PROTAC design tools
| Tool | Name | Description | Compute |
|---|---|---|---|
| Linker Enumeration | protac_linker_enumerate | Enumerate diverse linkers connecting a warhead to an E3 ligase ligand with control over length, flexibility, and chemical composition | GPU, 30-120s |
| Ternary Complex Modeling | protac_ternary_model | Model the three-body complex (POI-linker-E3 ligase) to assess geometric feasibility and protein-protein interface quality | GPU, 2-5min |
| Degradation Scoring | protac_degradation_score | Predict degradation efficiency metrics including DC50, Dmax, and hook effect risk from PROTAC structure and ternary complex data | CPU, 10-30s |
Capabilities
Linker enumeration
Generate libraries of chemically diverse linkers connecting a warhead (POI binder) to an E3 ligase ligand. The tool interface exposes a linker_type choice (peg, alkyl, click, mixed) and a length_range, but neither currently reaches the backend — see the note above. The tool generates 3D conformers for each linker and ranks them by synthetic accessibility and drug-likeness.
Ternary complex modeling
Model the three-body complex formed when a PROTAC brings a protein of interest (POI) into proximity with an E3 ubiquitin ligase. The tool takes the POI structure (PDB or predicted), the E3 ligase structure, and the full PROTAC molecule as input. It evaluates whether the linker geometry allows productive protein-protein contact and scores the interface by buried surface area, shape complementarity, and clash count.
Degradation scoring
Predict how effectively a PROTAC will degrade its target protein. The scoring model estimates DC50 (concentration for 50% degradation), Dmax (maximum degradation percentage), and flags compounds at risk of the hook effect (reduced degradation at high concentrations due to binary complex competition). Scores incorporate linker properties, ternary complex geometry, and physicochemical descriptors.
PROTAC design fundamentals
Mechanism of action
Key design parameters
| Parameter | Description | Typical Range |
|---|---|---|
| Linker length | Atoms in shortest path between warhead and E3 ligand | 4-16 atoms |
| Linker flexibility | Rotatable bonds in the linker | 2-10 |
| Molecular weight | Total PROTAC molecular weight | 700-1200 Da |
| DC50 | Concentration for 50% degradation | 1 nM - 10 uM |
| Dmax | Maximum degradation at saturation | 50-99% |
| Hook effect onset | Concentration where degradation decreases | >1 uM |
E3 ligase systems
| E3 Ligase | Ligand Class | Common Ligands | Advantages |
|---|---|---|---|
| CRBN (Cereblon) | IMiDs | Pomalidomide, lenalidomide, thalidomide | Broad tissue expression, well-characterized |
| VHL | Hydroxyproline | VH032, VHL-1 | High selectivity, no neosubstrate risk |
When to use each tool
| I want to... | Use this tool |
|---|---|
| Generate a library of linkers | protac_linker_enumerate |
| Test if my PROTAC geometry works | protac_ternary_model |
| Predict DC50 and Dmax | protac_degradation_score |
| Explore different linker chemistries | protac_linker_enumerate (linker_type, though see the note above — it isn't wired to the backend yet) |
| Check for hook effect risk | protac_degradation_score |
| Assess protein-protein interface quality | protac_ternary_model |
When to use other tools instead
| I want to... | Use this instead |
|---|---|
| Dock a small molecule inhibitor | DiffDock |
| Predict ADMET properties | ADMETlab 3 |
| Predict protein structure from sequence | ESMFold |
| Search for known PROTACs | ChEMBL or PubChem |
| Design a molecular glue (not bifunctional) | GenMol or REINVENT |
| Run retrosynthesis on a PROTAC | AiZynthFinder retrosynthesis |
Typical workflow
Citations
Sakamoto KM, et al. Protacs: chimeric molecules that target proteins to the Skp1-Cullin-F box complex for ubiquitination and degradation. Proc Natl Acad Sci USA. 2001;98(15):8554-8559. doi: 10.1073/pnas.141230798
Bondeson DP, et al. Catalytic in vivo protein knockdown by small-molecule PROTACs. Nat Chem Biol. 2015;11(8):611-617. doi: 10.1038/nchembio.1858
Churcher I. Protac-Induced Protein Degradation in Drug Discovery: Breaking the Rules or Just Making New Ones? J Med Chem. 2018;61(2):444-452. doi: 10.1021/acs.jmedchem.7b01272
Chat usage
Example prompts
- "Design a PROTAC targeting BRD4 with a CRBN recruiter and PEG linker"
- "Enumerate 50 linkers between 6 and 10 atoms for this warhead and VHL ligand"
- "Model the ternary complex for my BET degrader"
- "What DC50 and Dmax can I expect for these PROTAC candidates?"
- "Compare alkyl vs PEG linkers for my PROTAC series"
- "Check hook effect risk for my top 5 candidates"
Workflow chaining
- Feeds into: ADMET prediction (absorption, metabolism), retrosynthesis (synthetic route), compound passport (comprehensive assessment), molecular docking (binding validation)
- Accepts from: Molecular docking (warhead pose), protein structure prediction (POI structure), compound search (E3 ligand scaffolds), literature search (known PROTAC designs)
FAQ
Read next
CRISPR on the platform
Guide RNA design, off-target analysis, and CRISPR screen hit identification for genome editing experiments.
Oncology Analytics on the platform
Computational oncology tools for tumor mutational burden, microsatellite instability, and immunotherapy biomarker prediction.
Which tool?
Match a scientific goal to the tool that answers it.
Workflows
Guided pipelines and copy-and-adapt prompt recipes for common goals.
Was this page helpful?