If you want to find the gene-editing patents, you start with one classification code. The Cooperative Patent Classification — the shared scheme the USPTO and the European Patent Office use to sort patent documents by technology — places genetic engineering in subclass C12N. The USPTO's published CPC scheme gives C12N the title “MICROORGANISMS OR ENZYMES; COMPOSITIONS THEREOF; PROPAGATING, PRESERVING, OR MAINTAINING MICROORGANISMS; MUTATION OR GENETIC ENGINEERING; CULTURE MEDIA.” That one heading covers a wide band of biotechnology: living microorganisms, the enzymes derived from or acting on them, and — the part that matters most for the modern landscape — mutation and genetic engineering.

CPC is hierarchical, and the power of C12N for landscape analysis is in its subgroups. The code reads from broad to specific: section C (chemistry; metallurgy), class C12 (biochemistry; microbiology), subclass C12N, then numbered main groups and subgroups separated by a slash. So C12N 9 covers enzymes, with C12N 9/22 reaching nucleases — the endonucleases that cut DNA. C12N 15 covers mutation or genetic engineering and recombinant DNA technology, with C12N 15/11 reaching DNA or RNA fragments and their modification. These are not abstract labels; they are the buckets examiners assign and that analysts query to pull a clean set of patents in a given technology.

"C12N MICROORGANISMS OR ENZYMES; COMPOSITIONS THEREOF; PROPAGATING, PRESERVING, OR MAINTAINING MICROORGANISMS; MUTATION OR GENETIC ENGINEERING; CULTURE MEDIA"— USPTO, Cooperative Patent Classification scheme (C12N), source

Why C12N is the spine of the gene-editing landscape

The CRISPR estate lives in C12N, and the classifications on a real grant show why. The Broad Institute's U.S. Patent 11,578,312, “Engineering and optimization of systems, methods, enzymes and guide scaffolds of CAS9 orthologs and variants for sequence manipulation” (issued February 14, 2023), carries CPC symbols including C12N 9/22 — the nuclease code, capturing the Cas9 cutting enzyme — and C12N 15/11, capturing the nucleic-acid fragments such as the guide RNA. A landscape analyst who pulls every document classified in C12N 9/22 and C12N 15/11, faceted by assignee and filing year, is effectively drawing the map of the CRISPR-Cas patent field, because the classification follows the technology rather than the marketing name.

That is the analytical value of CPC. A keyword search for “CRISPR” misses documents that describe the same technology in different words and catches documents that merely mention the term in passing. A classification search anchored on C12N subgroups captures the patents an examiner determined are actually about the underlying invention — engineered nucleases, recombinant constructs, vectors — regardless of vocabulary. For freedom-to-operate work, white-space analysis, and filing-velocity trends, the C12N subgroup is a more reliable unit than a text query.

Reading a CPC symbol correctly is part of the skill. Take C12N 15/11: the “C12N” is the subclass, “15” is the main group covering mutation or genetic engineering, the slash separates the main group from the subgroup, and “11” is the subgroup reaching DNA or RNA fragments. The numbers after the slash are read as a hierarchy, not as a decimal — a longer subgroup number is more specific and sits beneath a shorter one. CPC also distinguishes invention-information symbols, which classify the inventive contribution of the document, from additional-information symbols, which capture other disclosed subject matter. A landscape analyst building a clean set generally weights the invention-information classifications, because those reflect what the examiner determined the document is actually about, while still scanning the additional symbols to catch cross-cutting subject matter that a narrow query would miss.

How C12N relates to the neighboring biotech classes

C12N does not stand alone. In the biologics and metabolic-drug estate, A61K (medicinal preparations and formulations) is where dosing and delivery claims cluster, and C07K (peptides, including antibodies) is where antibody composition claims sit. A single biotech patent is commonly classified in more than one of these, and the same Broad CRISPR grant above also carries C07K symbols for its peptide-related subject matter. Reading the full set of CPC symbols on a patent — not just the first one — tells you which technical neighborhoods the claims touch and, by extension, which portfolios and competitors a landscape map should include.

The maintenance of C12N also reflects how living the field is. CPC is revised on a rolling basis by the USPTO and the EPO, and subgroups are added or refined as technologies mature — the genetic-engineering portions of C12N 15 have expanded over time to accommodate recombinant techniques, gene therapy vectors, and editing systems that did not exist when the scheme was first laid out. For an analyst tracking filing velocity, this means a defensible time series sometimes has to account for reclassification: documents can be reclassified into newer subgroups, and a naive year-over-year count in a single code can understate or overstate a trend if the underlying classification boundaries moved. The discipline is to define the relevant C12N subgroup set deliberately, confirm it is stable across the window being measured, and then aggregate.

There is a further reason classification beats keyword for this field specifically: nomenclature in gene editing changes faster than the underlying technology. Terms like base editing, prime editing, and the various Cas orthologs entered the literature at different times, and a document filed before a term was coined will describe the same mechanism in older language. Because examiners assign C12N symbols based on the technical substance regardless of the words used, a classification-anchored search captures the early filings that a term-based search silently drops. For freedom-to-operate and white-space work, missing those early, differently-worded filings is exactly the kind of error that produces a falsely empty map.

For an IP-strategy reader, the practical takeaway is to treat C12N as the entry point for any gene-editing or genetic-engineering landscape, then descend to the specific subgroups — 9/22 for nucleases, 15/11 for nucleic-acid fragments, and the surrounding 15-series codes for recombinant and engineering subject matter. Aggregating real grants by those codes, rather than by keyword, is what produces a defensible picture of who is filing where, where the white space sits, and how filing velocity in the field is moving over time.

This article describes the classification scheme as published; it is a reference on how patents are organized, not legal advice, and the precise classification of any document is set by the examining office on the record.