THE BIOPHYSICAL DATA LAYER FOR DNA
Every AI model treats DNA as text. We compute what DNA actually does as a physical material.
30-SECOND QUICKSTART
pip install polymer-genomics
from polymer_genomics import PolymerClient
client = PolymerClient()
report = client.evaluate("ATGCGATCGATCGATCG" * 20)
print(report["flag_counts"]) # {"warnings": 0, "info": 2}
print(report["summary"]["gc_content"]) # 0.529OPEN ACCESS
No API key required. All endpoints are open during the research preview. Base URL: https://api.polymerbio.org
LIMITS & VERSIONS
API version:
0.2.0SDK version:
polymer-genomics 0.3.0Max sequence length (evaluate): 100,000 bp
Max region size (region query): 10 Mb
Max batch probes: 10,000
Max batch evaluate: 100 sequences
Max sequence retrieval: 100 kb
USE CASES
Synthetic Biology
Evaluate constructs before synthesis. Flag CpG islands that cause silencing, low-stability regions that misfold, and repeats that recombine.
client.evaluate(my_construct)AI Scientists
70 agent-composable MCP tools for biology research agents. Gene lookup, region queries, biophysical computation, cross-layer correlation — all in one server.
evaluate_design(sequence=...)Epigenetic Clocks
Mechanistic context for probe selection. 5 clock coefficient sets, 937K probe annotations, CpG context, and cross-array mapping.
client.clock_probes("horvath")DATA INVENTORY
29M
CpG sites
937K
methylation probes
63K
transcripts
54
GTEx tissues
70
MCP tools
30+
physical constants
CODE EXAMPLES
Python SDK
from polymer_genomics import PolymerClient
client = PolymerClient()
# Evaluate a construct for biophysical issues
report = client.evaluate(my_sequence)
for flag in report["flags"]:
print(f'{flag["code"]}: {flag["message"]}')
# Compare wild-type vs optimized design
delta = client.compare({
"wildtype": wt_sequence,
"optimized": opt_sequence,
})
print(delta["deltas_vs_reference"])
# Look up gene biophysics
tp53 = client.gene("hg38", "TP53")
expr = client.gene_expression("hg38", "TP53")
cost = client.gene_cost("hg38", "TP53")MCP Tools (Claude Code / AI Agents)
// claude_desktop_config.json
{
"mcpServers": {
"polymer-genomics": {
"command": "uvx",
"args": ["polymer-genomics-mcp"],
"env": {
"POLYMER_API_BASE": "https://api.polymerbio.org"
}
}
}
}
// Then in Claude Code:
// "Evaluate this construct for silencing risk"
// "What genes are near chr17:7668402-7687550?"
// "Compare these two promoter designs"REST API
curl -X POST https://api.polymerbio.org/v1/evaluate \
-H "Content-Type: application/json" \
-d '{"sequence": "ATGCGATCGATCG...", "name": "my_construct"}'
curl https://api.polymerbio.org/v1/genes/hg38/TP53
curl https://api.polymerbio.org/v1/biophysics/hg38/chr17:7668402-7669000