alterlab-adaptyv

name	alterlab-adaptyv
description	Submits and tracks protein-testing experiments on the Adaptyv Bio Foundry cloud lab (wet-lab validation), and optimizes protein sequences before submission with computational tools (NetSolP, SoluProt, SolubleMPNN, ESM). Use when designing proteins that need wet-lab validation - binding/affinity screening, expression testing, thermostability, or fluorescence assays - or when submitting experiments to the Foundry API, browsing the target catalog, tracking experiment status, retrieving results, or pre-screening sequences for solubility/expression. Triggers on "Adaptyv", "Foundry API", "cloud lab", "biolayer interferometry / BLI", "wet-lab validation". Part of the AlterLab Academic Skills suite.
license	MIT
allowed-tools	Read Write Edit Bash(python:*)
compatibility	Requires an Adaptyv Bio Foundry account and an ADAPTYV_API_KEY token for experiment submission; local protein-optimization steps (ESM, etc.) run via `uv run python` without a key.
metadata	{"skill-author":"AlterLab","version":"1.0.0"}

Adaptyv

Adaptyv Bio runs the Foundry cloud lab: submit protein sequences and a target, the lab runs the assay, and you retrieve experimental data (binding/affinity, thermostability, expression, fluorescence). The public Foundry API drives the full lifecycle programmatically. Turnaround is on the order of weeks; confirm the current estimate from the per-experiment quote rather than assuming a fixed number.

The exact request/response shapes evolve. This skill captures the verified API contract and conventions; for the authoritative spec see the OpenAPI doc at https://foundry-api-public.adaptyvbio.com/api/v1/openapi.json and https://docs.adaptyvbio.com.

Prefer the official tooling first

Adaptyv ships its own integrations - reach for them before hand-rolling requests:

• Official Python SDK — github.com/adaptyvbio/adaptyv-sdk (MIT). Decorator-based: wrap a design function with @lab.experiment(target=...); reads ADAPTYV_API_KEY / ADAPTYV_API_URL (and optional ADAPTYV_ORGANIZATION_ID) from the environment. Install from source (pip install -e . after cloning — no PyPI release confirmed; verify before pinning).
• Adaptyv's own Claude Code skills — github.com/adaptyvbio/protein-design-skills. Useful prior art for protein-design + Foundry workflows.

Use this skill's raw-requests recipes when the SDK is unavailable or you need fine control over the lifecycle.

Quick Start

Authentication Setup

1. Create a token in the Foundry portal: https://foundry.adaptyvbio.com/ → Organization → Settings → Tokens (pick a role: Member = read/write, Viewer = read-only; set an expiry). The token value is shown only once — copy it immediately.
2. Set it in your environment (never commit it):

export ADAPTYV_API_KEY="your_token_here"

Or put it in a gitignored .env:

ADAPTYV_API_KEY=your_token_here

Installation

If using the raw API directly:

uv pip install requests python-dotenv

Basic Usage

The API uses a draft → submit flow: create an experiment (it starts as a draft), then submit it. sequences is a {label: amino_acid_string} map (multi-chain constructs join chains with a colon, e.g. "heavy:light").

import os
import requests
from dotenv import load_dotenv

load_dotenv()

api_key = os.getenv("ADAPTYV_API_KEY")
base_url = "https://foundry-api-public.adaptyvbio.com/api/v1"
headers = {
    "Authorization": f"Bearer {api_key}",
    "Content-Type": "application/json",
}

# 1. Create a draft experiment
resp = requests.post(
    f"{base_url}/experiments",
    headers=headers,
    json={
        "name": "mini-binder round 1",
        "experiment_spec": {
            "experiment_type": "affinity",   # screening|affinity|thermostability|fluorescence|expression
            "method": "bli",                 # bli|spr (for binding-type assays)
            "target_id": "<target uuid from GET /targets>",
            "sequences": {
                "design_a": "MKVLWALLGLLGAA...",
                "design_b": "MATGVLWALLG...",
            },
        },
    },
)
resp.raise_for_status()
experiment_id = resp.json()["experiment_id"]

# 2. Submit it to the lab (after reviewing the quote — see reference/api_reference.md)
requests.post(f"{base_url}/experiments/{experiment_id}/submit", headers=headers).raise_for_status()

Available Experiment Types

Foundry supports these experiment_type values:

• screening - Binding detection via biolayer interferometry (BLI) or SPR. Requires a target.
• affinity - Kinetic constants (KD, kon, koff) by BLI/SPR. Requires a target.
• thermostability - Melting temperature (Tm) via DSF. No target required.
• fluorescence - Fluorescence intensity. No target required.
• expression - Protein yield quantification. No target required.

See reference/experiments.md for detailed information on each assay and its outputs.

Protein Sequence Optimization

Before submitting sequences, optimize them for better expression and stability:

Common issues to address:

• Unpaired cysteines that create unwanted disulfides
• Excessive hydrophobic regions causing aggregation
• Poor solubility predictions

Recommended tools:

• NetSolP / SoluProt - Initial solubility filtering (both are web services, not pip packages)
• SolubleMPNN - Solubility-biased sequence redesign (a weight set within the ProteinMPNN / LigandMPNN family)
• ESM (fair-esm) - Sequence likelihood / naturalness scoring
• ipTM (AlphaFold-Multimer / ColabFold) - Interface stability for binder designs
• pSAE - Solvent-accessible hydrophobic exposure, from a predicted/known structure

See reference/protein_optimization.md for detailed optimization workflows and tool usage.

API Reference

For complete API documentation including all endpoints, request/response formats, and authentication details, see reference/api_reference.md.

Examples

For concrete code examples covering common use cases (experiment submission, status tracking, result retrieval, batch processing), see reference/examples.md.

Important Notes

• The Foundry API is public but still evolving — treat the OpenAPI doc (/api/v1/openapi.json) as the source of truth and verify field names before relying on them.
• Submission is two-step: create a draft, review the cost quote, then POST .../submit. Nothing is charged until you confirm the quote.
• affinity/screening require a target_id from the catalog (GET /targets); thermostability, fluorescence, and expression do not.
• Turnaround is multiple weeks — read the estimate from the experiment/quote rather than assuming a fixed number.
• Support and docs: [e-mail] / https://docs.adaptyvbio.com.
• Suitable for high-throughput AI-driven protein design workflows (closed-loop design → test → learn).