---
title: PROTAC design on the platform
description: >-
  Targeted protein degradation tools for PROTAC linker enumeration, ternary
  complex modeling, and degradation efficiency prediction.
tags:
  - protac
---

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](/docs/targets-modalities) 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

 I[Intent detection]
    I -->|Linker design| L[PROTAC Linker Enumerate]
    I -->|Complex modeling| T[PROTAC Ternary Model]
    I -->|Degradation prediction| D[PROTAC Degradation Score]
    L & T & D --> C["GPU compute (Modal, 10 s - 5 min)"]
    C --> Res[Structured results with 3D coordinates / scores]`}
  caption="A PROTAC request is routed by intent to linker design, complex modeling, or degradation scoring, then run on GPU."
/>

## 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

 B[Ternary complex formation<br/>POI-PROTAC-E3]
    B --> C[E3 ligase ubiquitinates POI]
    C --> D[Ubiquitinated POI recognized<br/>by 26S proteasome]
    D --> E[POI degraded into peptide fragments]
    E --> F[PROTAC released]
    F -->|catalytic cycle repeats| A`}
  caption="A PROTAC acts catalytically: it brings the target and an E3 ligase together, is released after degradation, and repeats the cycle."
/>

### 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 | &gt;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](/docs/diffdock) |
| Predict ADMET properties | [ADMETlab 3](/docs/admetlab3) |
| Predict protein structure from sequence | [ESMFold](/docs/esmfold) |
| Search for known PROTACs | [ChEMBL](/docs/chembl) or [PubChem](/docs/pubchem) |
| Design a molecular glue (not bifunctional) | [GenMol](/docs/genmol) or [REINVENT](/docs/reinvent) |
| Run retrosynthesis on a PROTAC | [AiZynthFinder](/docs/aizynthfinder) retrosynthesis |

## Typical workflow

```
1. Select warhead (from docking or literature)
2. Select E3 ligase system (CRBN or VHL)
3. Enumerate linkers (protac_linker_enumerate)
4. Filter by drug-likeness and synthetic accessibility
5. Model ternary complexes for top candidates (protac_ternary_model)
6. Score degradation efficiency (protac_degradation_score)
7. Select leads for synthesis
```

## 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](https://doi.org/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](https://doi.org/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](https://doi.org/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

### What is a PROTAC?

  A PROTAC (Proteolysis-Targeting Chimera) is a bifunctional molecule with a warhead that binds the protein of interest (POI), a ligand that recruits an E3 ubiquitin ligase, and a linker connecting them. PROTACs induce proximity between the POI and E3 ligase, leading to ubiquitination and proteasomal degradation of the target protein.

### Which E3 ligases are supported?

  For linker enumeration, provide any E3 ligand as `e3_binder_smiles` (SMILES) — CRBN (thalidomide/pomalidomide/lenalidomide-derived) and VHL (VH032/VHL-1-derived) are the common choices. `protac_degradation_score` accepts an `e3_ligase` selector (`CRBN`, `VHL`, `IAP`, `MDM2`), but per the note above the backend doesn't currently read that field, so it has no effect on the score.

### What is the difference between DC50 and Dmax?

  DC50 is the concentration at which 50% of the target protein is degraded (analogous to IC50 for inhibitors). Dmax is the maximum percentage of target protein that can be degraded at saturating concentrations. An ideal PROTAC has a low DC50 (potent) and high Dmax (near-complete degradation).

### What is the hook effect?

  The hook effect occurs at high PROTAC concentrations when binary complexes (PROTAC-POI or PROTAC-E3) outcompete the productive ternary complex (POI-PROTAC-E3). This causes reduced degradation at high doses, producing a bell-shaped dose-response curve. The degradation scoring tool flags compounds at risk of the hook effect.

## Read next