diffdock-nim

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

DiffDock NIM

Predict protein-ligand binding poses with blind docking. Use this

SKILL.md

for first-pass hosted/local usage; load supplemental files only when needed:

```
references/api.md
```
: exact hosted/local endpoints, schemas, Docker flags.
```
references/science.md
```
: docking use cases, limits, and handoffs.
```
references/parameters.md
```
: ligand formats, pose counts, diffusion controls.
```
references/validation.md
```
: receptor, ligand, pose, and confidence checks.
```
references/examples.md
```
: compact hosted/local and pose-saving patterns.

通过盲对接预测蛋白-配体结合构象。本

SKILL.md

适用于首次使用托管/本地部署场景；仅在需要时加载补充文件：

```
references/api.md
```
：托管/本地端点、数据结构、Docker参数的详细说明。
```
references/science.md
```
：对接使用场景、限制条件及交接说明。
```
references/parameters.md
```
：配体格式、构象数量、扩散控制参数。
```
references/validation.md
```
：受体、配体、构象及置信度校验说明。
```
references/examples.md
```
：托管/本地部署及构象保存的简洁示例。

Choose Mode

选择运行模式

Ask only when context is unclear:

Hosted NVIDIA API or local Docker NIM?

Hosted:

https://health.api.nvidia.com/v1/biology/mit/diffdock

Local:

http://localhost:8000/molecular-docking/diffdock/generate

The hosted and local paths differ. Local has no

/v1/

prefix and uses the

/molecular-docking/

route. Hosted requests use

Authorization: Bearer $NGC_API_KEY

. Supported local Docker startup uses

NGC_API_KEY

(or

NVIDIA_API_KEY

via the preflight) for registry login, entitlement checks, and first-run model downloads; pass it into the container with

-e NGC_API_KEY

. Local inference requests use no auth header after readiness. Warm-cache key-free startup varies by image/version and should not be assumed.

仅在上下文不明确时询问：

使用NVIDIA托管API还是本地Docker NIM？

托管版：

https://health.api.nvidia.com/v1/biology/mit/diffdock

本地版：

http://localhost:8000/molecular-docking/diffdock/generate

托管版与本地版的路径不同。本地版没有

/v1/

前缀，使用

/molecular-docking/

路由。托管版请求需使用

Authorization: Bearer $NGC_API_KEY

。本地Docker启动需使用

NGC_API_KEY

（或通过预检流程使用

NVIDIA_API_KEY

）进行镜像仓库登录、权限校验及首次运行时的模型下载；需通过

-e NGC_API_KEY

将其传入容器。本地推理请求在服务就绪后无需认证头。无密钥的预热缓存启动方式因镜像/版本而异，请勿默认依赖该方式。

Local Docker

本地Docker部署

For the exact local preflight (

.env

load,

NVIDIA_API_KEY

fallback,

LOCAL_NIM_CACHE

NVIDIA_VISIBLE_DEVICES=0

--shm-size=2G

, both

--ulimit

flags,

docker login

, and the

docker run

for

nvcr.io/nim/mit/diffdock:2.2.0

), copy the command block in

references/api.md

under Docker Reference verbatim.

Readiness:

bash

until curl -sf http://localhost:8000/v1/health/ready; do sleep 5; done

如需完整的本地预检流程（加载

.env

文件、

NVIDIA_API_KEY

fallback、

LOCAL_NIM_CACHE

、

NVIDIA_VISIBLE_DEVICES=0

、

--shm-size=2G

、两个

--ulimit

参数、

docker login

以及运行

nvcr.io/nim/mit/diffdock:2.2.0

的

docker run

命令），请直接复制

references/api.md

中Docker参考下的命令块。

服务就绪校验：

bash

until curl -sf http://localhost:8000/v1/health/ready; do sleep 5; done

Prepare Inputs

准备输入数据

Protein receptor must be ATOM records only. Strip headers, water, and HETATM.

python

from pathlib import Path
raw_pdb = Path("protein.pdb").read_text()
protein = "\n".join(line for line in raw_pdb.splitlines() if line.startswith("ATOM"))
if not protein:
    raise ValueError("protein.pdb has no ATOM records")

Ligand options:

SMILES:

ligand = "CC(=O)OC1=CC=CC=C1C(=O)O"

;

ligand_file_type = "txt"

SDF:

ligand = Path("ligand.sdf").read_text()

;

ligand_file_type = "sdf"

MOL2:
```
ligand_file_type = "mol2"
```
.

Do not use

"smiles"

ligand_file_type

; SMILES is

"txt"

蛋白受体必须仅包含ATOM记录。移除头部信息、水分子及HETATM记录。

python

from pathlib import Path
raw_pdb = Path("protein.pdb").read_text()
protein = "\n".join(line for line in raw_pdb.splitlines() if line.startswith("ATOM"))
if not protein:
    raise ValueError("protein.pdb has no ATOM records")

配体选项：

SMILES格式：

ligand = "CC(=O)OC1=CC=CC=C1C(=O)O"

；

ligand_file_type = "txt"

。

SDF格式：

ligand = Path("ligand.sdf").read_text()

；

ligand_file_type = "sdf"

。

MOL2格式：
```
ligand_file_type = "mol2"
```
。

请勿将

"smiles"

作为

ligand_file_type

的值；SMILES格式对应的是

"txt"

。

Request Pattern

请求示例

python

import os
import requests

HOSTED = True
url = (
    "https://health.api.nvidia.com/v1/biology/mit/diffdock"
    if HOSTED else "http://localhost:8000/molecular-docking/diffdock/generate"
)
headers = {"Content-Type": "application/json"}
if HOSTED:
    headers["Authorization"] = f"Bearer {os.getenv('NGC_API_KEY')}"

payload = {
    "protein": protein,
    "ligand": ligand,
    "ligand_file_type": ligand_file_type,
    "num_poses": 10,
    "time_divisions": 20,
    "steps": 18,
    "save_trajectory": False,
}
response = requests.post(url, headers=headers, json=payload, timeout=300)
response.raise_for_status()
result = response.json()

python

import os
import requests

HOSTED = True
url = (
    "https://health.api.nvidia.com/v1/biology/mit/diffdock"
    if HOSTED else "http://localhost:8000/molecular-docking/diffdock/generate"
)
headers = {"Content-Type": "application/json"}
if HOSTED:
    headers["Authorization"] = f"Bearer {os.getenv('NGC_API_KEY')}"

payload = {
    "protein": protein,
    "ligand": ligand,
    "ligand_file_type": ligand_file_type,
    "num_poses": 10,
    "time_divisions": 20,
    "steps": 18,
    "save_trajectory": False,
}
response = requests.post(url, headers=headers, json=payload, timeout=300)
response.raise_for_status()
result = response.json()

Save And Report Output

保存并报告输出结果

ligand_positions

and

position_confidence

are parallel ranked lists.

position_confidence[0]

is the rank-1 pose confidence.

Save the ranked pose SDFs using the snippet in

references/examples.md

under Save Ranked Poses.

View pose SDF files with the receptor in PyMOL, ChimeraX, or UCSF Chimera. For pose sanity checks and confidence caveats, read

references/validation.md

ligand_positions

与

position_confidence

是相互对应的排序后列表。

position_confidence[0]

为排名第一的构象置信度。

请使用

references/examples.md

中保存排序后构象下的代码片段保存排序后的构象SDF文件。

可在PyMOL、ChimeraX或UCSF Chimera中结合受体查看构象SDF文件。如需了解构象合理性校验及置信度相关注意事项，请阅读

references/validation.md

。

Limits And Troubleshooting

限制条件与故障排查

Max
```
num_poses
```
: 100. Max
```
time_divisions
```
: 20. Max
```
steps
```
: 18.
Single GPU; local minimum is about 24 GB VRAM.
```
422
```
: invalid
```
ligand_file_type
```
, invalid SMILES/SDF, or no ATOM records.
Empty poses: validate receptor ATOM records and ligand parseability.
Local URL 404 usually means the wrong hosted path or an accidental
```
/v1/
```
.

```
num_poses
```
最大值：100。
```
time_divisions
```
最大值：20。
```
steps
```
最大值：18。
仅支持单GPU；本地部署最低需要约24GB显存。
返回
```
422
```
错误：
```
ligand_file_type
```
无效、SMILES/SDF格式无效或无ATOM记录。
构象为空：校验受体ATOM记录及配体可解析性。
本地URL返回404错误：通常是使用了错误的托管版路径或误加了
```
/v1/
```
前缀。