Building the Family Tree for a Biological Brand

Open the SBI genealogy explorer and you'll see a network of colored nodes and connecting lines — a visual map of how cannabis strains relate to each other across generations. Behind that interface is a database of over 27,000 documented genetic connections, making it one of the most comprehensive cannabis family trees ever assembled.

This is the story of how we built it.

Starting with the Historical Record

Cannabis genealogy is an unusual data problem. Unlike agricultural crops that have been systematically documented by universities and government agencies for a century, cannabis breeding has been largely underground. The historical record exists in:

• Breeder interviews and oral histories — conversations with the people who made the crosses, many of whom are now in their 60s and 70s.
• Forum archives — two decades of cannabis community forums where breeders discussed their work publicly.
• Seed catalogs — historical listings from Dutch and Spanish seed banks that documented lineage information.
• Published research — the small but growing body of academic work on cannabis genetics and chemotaxonomy.

We started by aggregating these sources into a structured format. Every claimed parent-child relationship was documented with its source and confidence level. Where sources disagreed (and they frequently do), we recorded all versions rather than picking a winner.

The Data Model

The genealogy database uses a directed acyclic graph (DAG) — the same data structure used in version control systems like Git. Each strain is a node. Each breeding relationship is a directed edge from parent to child.

A typical entry contains:

• Strain identity — name, aliases, breeder, year of creation (if known)
• Genetic classification — indica/sativa/hybrid percentage, based on documented parentage rather than subjective assessment
• Parent relationships — mother and father (for sexual crosses), or single parent for feminized/selfed lines (S1, reversed)
• Breeding method — F1, F2, BX (backcross), S1 (selfed), polyhybrid, etc.
• Confidence score — how well-documented is this relationship? Primary source (breeder confirmation), secondary source (reputable publication), or community consensus.

The DAG structure means you can traverse the tree in both directions: trace a strain's ancestry back through its parents, grandparents, and great-grandparents, or look forward to see all documented descendants of a foundational line.

The Depth Problem

Cannabis genealogy gets sparse fast. For modern cultivars — anything created in the last 10-15 years — we can usually trace parentage back 2-3 generations with high confidence. But beyond that, documentation becomes unreliable.

The classic example: OG Kush. Despite being one of the most influential cultivars in modern cannabis, its exact parentage remains debated. Various sources claim it's a Chemdawg phenotype, a Hindu Kush cross, or something else entirely. Our database records all claimed lineages with appropriate confidence scores.

For foundational genetics like Skunk #1, Northern Lights, and the original Haze, we rely on breeder accounts that are decades old and sometimes contradictory. We've documented these discrepancies openly — the database shows what's known, what's disputed, and what's uncertain.

Scale and Growth

The 27,000 connections in our database break down roughly as follows:

• ~4,000 unique cultivars — distinct named strains with at least one documented genetic relationship
• ~700 cultivars with full SBI verification — strains listed on the platform with breeder-confirmed lineage
• ~300 foundational genetics — landraces and first-generation hybrids that appear as ancestors of multiple modern lines
• ~27,000 parent-child edges — the connections between all of these, including multiple claimed parentages for disputed lineages

The database grows in two ways: new strains added by verified breeders on the platform, and historical research that fills in gaps in older lineages.

What the Family Tree Reveals

With enough data, patterns emerge. Some of what we've found:

Genetic bottlenecks are real. Approximately 80% of modern commercial cultivars trace back to fewer than 50 foundational genetics. The diversity you see on dispensary shelves is, in many cases, a recombination of a surprisingly narrow genetic base.

Naming is chaos. The same genetics circulate under different names in different markets. Our database has documented over 200 cases where the same cultivar (confirmed by matching lineage and breeder origin) is sold under two or more names.

Some breeders are disproportionately influential. A small number of foundational breeders — perhaps 20-30 individuals — are responsible for genetics that appear in the ancestry of thousands of modern strains. Their work is the bedrock of the industry, yet most consumers have never heard their names.

An Open Resource

The genealogy database is a public resource. Per the SBI Charter (Article VI), the genetic lineage map belongs to everyone. Individual transaction data remains private, but the family tree itself — the documented relationships between cultivars — is openly accessible.

We believe this data has value beyond seed sales. For researchers studying cannabis chemotaxonomy, for breeders planning crosses, for regulators trying to establish cultivar standards — a comprehensive, structured, publicly available genetic database is infrastructure that the entire industry needs.

Explore the family tree at /genealogy. If you have lineage information that could improve the database — especially for older or regional cultivars — we want to hear from you.