Tools · Computational Biology

RFAntibody

Antibody and TCR design with RFdiffusion, ProteinMPNN, and RF2 (stages or full pipeline)

On this page

Prices, workflows, and method papers Open in the app

Run it from the API

Submit with Submit a job and the job_type below. Price it first with Estimate job reservation cost: submitting reserves that amount from your wallet, and the charge settles at the actual runtime.

RFAntibody Fine-tuning `rfantibody-finetune`

Job type: rfantibody-finetune
Hardware: t4l4a10l40sa100-40gb (default)a100-80gbh100h200b200
Typical runtime: 2.3 h on Nvidia A100, 40 GB

Payload

Payload fields
Field	Type	Description
`model_target`required	`string`	One of: `"rfdiffusion", "proteinmpnn", "rf2"`
`training_data`required	`string`	PDB text for rfdiffusion/rf2, or FASTA sequences for proteinmpnn. Limits: `min length 1`

Example

from cognichem_client import CogniChem

client = CogniChem.from_env()  # reads COGNICHEM_API_KEY
payload = {
    "model_target": "rfdiffusion",
    "training_data": "ATOM fake training row\n",
}

estimate = client.jobs.estimate(job_type="rfantibody-finetune", payload=payload, resource="a100-40gb")
print(f"Reserves ${estimate.cost:.2f}")

job = client.jobs.submit(
    job_name="my-rfantibody-finetune-run",
    job_type="rfantibody-finetune",
    payload=payload,
    resource="a100-40gb",
)
status = client.jobs.wait(job.process_id)
if status.status == "completed":
    client.jobs.result(job.process_id, save_path=".")

curl -X POST "https://api.cognichem.com/api/v1/jobs/submit" \
  -H "X-Api-Key: $COGNICHEM_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
  "job_name": "my-rfantibody-finetune-run",
  "job_type": "rfantibody-finetune",
  "payload": {
    "model_target": "rfdiffusion",
    "training_data": "ATOM fake training row\n"
  },
  "resource": "a100-40gb"
}'

{
  "model_target": "rfdiffusion",
  "training_data": "ATOM fake training row\n"
}

Sample data from the job catalog; long values are shortened here. Each job_name must be unique among your jobs.

Workflow inputs

None

Workflow outputs

ArchiveZIP

RFAntibody `rfantibody-pipeline`

Job type: rfantibody-pipeline
Hardware: t4l4a10l40sa100-40gb (default)a100-80gbh100h200b200
Typical runtime: 2 h on Nvidia A100, 40 GB

Payload

Payload fields
Field	Type	Description
`target_pdb`required	`string`	Target antigen PDB content with ATOM/HETATM records. Limits: `min length 1`
`hotspot_residues[]`required	`string[]`	Limits: `min items 1`
`framework`	`string`	Default: `"vhh"`One of: `"vhh", "scfv"`
`num_designs`	`integer`	Default: `10`Limits: `≥ 1`
`diffuser_T`	`integer`	Default: `50`Limits: `≥ 15`
`final_step`	`integer`	Default: `1`Limits: `≥ 1`
`mpnn_temperature`	`number`	Default: `0.1`Limits: `> 0`
`mpnn_seqs_per_struct`	`integer`	Default: `1`Limits: `≥ 1`
`rf2_pae_threshold`	`number`	Default: `10`Limits: `> 0`
`rf2_rmsd_threshold`	`number`	Default: `2`Limits: `> 0`
`tcr_mode`	`boolean`	Forced true for rfantibody-tcr-predict by the API router.

Example

from cognichem_client import CogniChem

client = CogniChem.from_env()  # reads COGNICHEM_API_KEY
payload = {
    "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
    "hotspot_residues": ["A:1", "A:2"],
    "num_designs": 1,
}

estimate = client.jobs.estimate(job_type="rfantibody-pipeline", payload=payload, resource="a100-40gb")
print(f"Reserves ${estimate.cost:.2f}")

job = client.jobs.submit(
    job_name="my-rfantibody-pipeline-run",
    job_type="rfantibody-pipeline",
    payload=payload,
    resource="a100-40gb",
)
status = client.jobs.wait(job.process_id)
if status.status == "completed":
    client.jobs.result(job.process_id, save_path=".")

curl -X POST "https://api.cognichem.com/api/v1/jobs/submit" \
  -H "X-Api-Key: $COGNICHEM_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
  "job_name": "my-rfantibody-pipeline-run",
  "job_type": "rfantibody-pipeline",
  "payload": {
    "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
    "hotspot_residues": [
      "A:1",
      "A:2"
    ],
    "num_designs": 1
  },
  "resource": "a100-40gb"
}'

{
  "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
  "hotspot_residues": [
    "A:1",
    "A:2"
  ],
  "num_designs": 1
}

Sample data from the job catalog; long values are shortened here. Each job_name must be unique among your jobs.

Workflow inputs

Protein structurePDB

Workflow outputs

ArchiveZIP
StructuresPDB

RFAntibody ProteinMPNN `rfantibody-proteinmpnn`

Job type: rfantibody-proteinmpnn
Hardware: t4l4a10 (default)l40sa100-40gba100-80gbh100h200b200
Typical runtime: 30 min on Nvidia A10

Payload

Payload fields
Field	Type	Description
`target_pdb`required	`string`	Target antigen PDB content with ATOM/HETATM records. Limits: `min length 1`
`hotspot_residues[]`required	`string[]`	Limits: `min items 1`
`framework`	`string`	Default: `"vhh"`One of: `"vhh", "scfv"`
`num_designs`	`integer`	Default: `10`Limits: `≥ 1`
`diffuser_T`	`integer`	Default: `50`Limits: `≥ 15`
`final_step`	`integer`	Default: `1`Limits: `≥ 1`
`mpnn_temperature`	`number`	Default: `0.1`Limits: `> 0`
`mpnn_seqs_per_struct`	`integer`	Default: `1`Limits: `≥ 1`
`rf2_pae_threshold`	`number`	Default: `10`Limits: `> 0`
`rf2_rmsd_threshold`	`number`	Default: `2`Limits: `> 0`
`tcr_mode`	`boolean`	Forced true for rfantibody-tcr-predict by the API router.

Example

from cognichem_client import CogniChem

client = CogniChem.from_env()  # reads COGNICHEM_API_KEY
payload = {
    "target_pdb": "ATOM      1  N   GLU H   1      23.793  -8.718 -21.757  1.00… (36,398 characters)",
    "hotspot_residues": ["T:129", "T:130"],
    "num_designs": 1,
}

estimate = client.jobs.estimate(job_type="rfantibody-proteinmpnn", payload=payload, resource="a10")
print(f"Reserves ${estimate.cost:.2f}")

job = client.jobs.submit(
    job_name="my-rfantibody-proteinmpnn-run",
    job_type="rfantibody-proteinmpnn",
    payload=payload,
    resource="a10",
)
status = client.jobs.wait(job.process_id)
if status.status == "completed":
    client.jobs.result(job.process_id, save_path=".")

curl -X POST "https://api.cognichem.com/api/v1/jobs/submit" \
  -H "X-Api-Key: $COGNICHEM_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
  "job_name": "my-rfantibody-proteinmpnn-run",
  "job_type": "rfantibody-proteinmpnn",
  "payload": {
    "target_pdb": "ATOM      1  N   GLU H   1      23.793  -8.718 -21.757  1.00… (36,398 characters)",
    "hotspot_residues": [
      "T:129",
      "T:130"
    ],
    "num_designs": 1
  },
  "resource": "a10"
}'

{
  "target_pdb": "ATOM      1  N   GLU H   1      23.793  -8.718 -21.757  1.00… (36,398 characters)",
  "hotspot_residues": [
    "T:129",
    "T:130"
  ],
  "num_designs": 1
}

Sample data from the job catalog; long values are shortened here. Each job_name must be unique among your jobs.

Workflow inputs

Protein structurePDB

Workflow outputs

ArchiveZIP
StructuresPDB

RFAntibody RFdiffusion `rfantibody-rfdiffusion`

Job type: rfantibody-rfdiffusion
Hardware: t4l4a10l40sa100-40gb (default)a100-80gbh100h200b200
Typical runtime: 1.5 h on Nvidia A100, 40 GB

Payload

Payload fields
Field	Type	Description
`target_pdb`required	`string`	Target antigen PDB content with ATOM/HETATM records. Limits: `min length 1`
`hotspot_residues[]`required	`string[]`	Limits: `min items 1`
`framework`	`string`	Default: `"vhh"`One of: `"vhh", "scfv"`
`num_designs`	`integer`	Default: `10`Limits: `≥ 1`
`diffuser_T`	`integer`	Default: `50`Limits: `≥ 15`
`final_step`	`integer`	Default: `1`Limits: `≥ 1`
`mpnn_temperature`	`number`	Default: `0.1`Limits: `> 0`
`mpnn_seqs_per_struct`	`integer`	Default: `1`Limits: `≥ 1`
`rf2_pae_threshold`	`number`	Default: `10`Limits: `> 0`
`rf2_rmsd_threshold`	`number`	Default: `2`Limits: `> 0`
`tcr_mode`	`boolean`	Forced true for rfantibody-tcr-predict by the API router.

Example

from cognichem_client import CogniChem

client = CogniChem.from_env()  # reads COGNICHEM_API_KEY
payload = {
    "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
    "hotspot_residues": ["A:1", "A:2"],
    "num_designs": 1,
}

estimate = client.jobs.estimate(job_type="rfantibody-rfdiffusion", payload=payload, resource="a100-40gb")
print(f"Reserves ${estimate.cost:.2f}")

job = client.jobs.submit(
    job_name="my-rfantibody-rfdiffusion-run",
    job_type="rfantibody-rfdiffusion",
    payload=payload,
    resource="a100-40gb",
)
status = client.jobs.wait(job.process_id)
if status.status == "completed":
    client.jobs.result(job.process_id, save_path=".")

curl -X POST "https://api.cognichem.com/api/v1/jobs/submit" \
  -H "X-Api-Key: $COGNICHEM_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
  "job_name": "my-rfantibody-rfdiffusion-run",
  "job_type": "rfantibody-rfdiffusion",
  "payload": {
    "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
    "hotspot_residues": [
      "A:1",
      "A:2"
    ],
    "num_designs": 1
  },
  "resource": "a100-40gb"
}'

{
  "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
  "hotspot_residues": [
    "A:1",
    "A:2"
  ],
  "num_designs": 1
}

Sample data from the job catalog; long values are shortened here. Each job_name must be unique among your jobs.

Workflow inputs

Protein structurePDB

Workflow outputs

ArchiveZIP
StructuresPDB

RFAntibody TCR Predictor `rfantibody-tcr-predict`

Job type: rfantibody-tcr-predict
Hardware: t4l4a10l40sa100-40gb (default)a100-80gbh100h200b200
Typical runtime: 1 h on Nvidia A100, 40 GB

Payload

Payload fields
Field	Type	Description
`target_pdb`required	`string`	Target antigen PDB content with ATOM/HETATM records. Limits: `min length 1`
`hotspot_residues[]`required	`string[]`	Limits: `min items 1`
`framework`	`string`	Default: `"vhh"`One of: `"vhh", "scfv"`
`num_designs`	`integer`	Default: `10`Limits: `≥ 1`
`diffuser_T`	`integer`	Default: `50`Limits: `≥ 15`
`final_step`	`integer`	Default: `1`Limits: `≥ 1`
`mpnn_temperature`	`number`	Default: `0.1`Limits: `> 0`
`mpnn_seqs_per_struct`	`integer`	Default: `1`Limits: `≥ 1`
`rf2_pae_threshold`	`number`	Default: `10`Limits: `> 0`
`rf2_rmsd_threshold`	`number`	Default: `2`Limits: `> 0`
`tcr_mode`	`boolean`	Forced true for rfantibody-tcr-predict by the API router.

Example

from cognichem_client import CogniChem

client = CogniChem.from_env()  # reads COGNICHEM_API_KEY
payload = {
    "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
    "hotspot_residues": ["A:1", "A:2"],
    "num_designs": 1,
    "tcr_mode": True,
}

estimate = client.jobs.estimate(job_type="rfantibody-tcr-predict", payload=payload, resource="a100-40gb")
print(f"Reserves ${estimate.cost:.2f}")

job = client.jobs.submit(
    job_name="my-rfantibody-tcr-predict-run",
    job_type="rfantibody-tcr-predict",
    payload=payload,
    resource="a100-40gb",
)
status = client.jobs.wait(job.process_id)
if status.status == "completed":
    client.jobs.result(job.process_id, save_path=".")

curl -X POST "https://api.cognichem.com/api/v1/jobs/submit" \
  -H "X-Api-Key: $COGNICHEM_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
  "job_name": "my-rfantibody-tcr-predict-run",
  "job_type": "rfantibody-tcr-predict",
  "payload": {
    "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
    "hotspot_residues": [
      "A:1",
      "A:2"
    ],
    "num_designs": 1,
    "tcr_mode": true
  },
  "resource": "a100-40gb"
}'

{
  "target_pdb": "REMARK  Minimal alanine tripeptide for unit testing.\nREMARK … (2,166 characters)",
  "hotspot_residues": [
    "A:1",
    "A:2"
  ],
  "num_designs": 1,
  "tcr_mode": true
}

Sample data from the job catalog; long values are shortened here. Each job_name must be unique among your jobs.

Workflow inputs

Protein structurePDB

Workflow outputs

ArchiveZIP
StructuresPDB

Run it from the API

RFAntibody Fine-tuning rfantibody-finetune

RFAntibody rfantibody-pipeline

RFAntibody ProteinMPNN rfantibody-proteinmpnn

RFAntibody RFdiffusion rfantibody-rfdiffusion

RFAntibody TCR Predictor rfantibody-tcr-predict

RFAntibody Fine-tuning `rfantibody-finetune`

RFAntibody `rfantibody-pipeline`

RFAntibody ProteinMPNN `rfantibody-proteinmpnn`

RFAntibody RFdiffusion `rfantibody-rfdiffusion`

RFAntibody TCR Predictor `rfantibody-tcr-predict`