A Guide to the Systematic Nomenclature of Lipids: Structure, Codes, and Shorthand Rules

The systematic nomenclature of lipids is the standardized naming system, now principally defined by the LIPID MAPS classification and the IUPAC-IUBMB recommendations, that assigns every lipid a two-part identity: a category-based abbreviation (such as PC 34:1 or TAG 52:3) and an exact structural name that specifies headgroup, backbone, chain lengths, double-bond positions, and stereochemistry. The shorthand tells you the bulk composition; the systematic name removes all ambiguity. Together they allow a lipid extracted from a cell membrane, a plasma sample, or a food matrix to be described in a form another laboratory can reproduce without ever seeing the molecule. Getting the nomenclature right is not bookkeeping, because PC 34:1 can denote dozens of distinct molecular species.

How lipid names encode molecular structure

Every systematic lipid name answers four questions in a fixed order: what kind of lipid is it, what is the backbone, which chains are attached, and where are the double bonds. The chain description carries most of the information. A fatty acid named 18:1(9Z) has 18 carbons, one double bond, and that bond begins at carbon 9 counting from the carboxyl carbon, with cis (Z) geometry. Shift the double bond to carbon 11 and you have a different molecule, cis-vaccenic acid, with different physical behavior.

Positional information is equally strict. Glycerolipids number the glycerol carbons sn-1, sn-2, and sn-3, where sn stands for stereospecific numbering. The hydroxyl at sn-2 points in the opposite direction from those at sn-1 and sn-3, which is why enzymes such as phospholipase A2 can strip the sn-2 acyl chain with near-total selectivity. A name like 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine pins down exactly which chain sits where. Written as PC 16:0/18:1, the slash carries that positional promise; written as PC 34:1, it does not.

Sphingolipids use their own building blocks. Sphingosine is (2S,3R,4E)-2-aminooctadec-4-ene-1,3-diol, and ceramides attach a fatty acid to its nitrogen via an amide bond. The N- prefix marks that amide linkage, so N-palmitoyl-sphingosine and palmitoyl-sphingosine mean the same thing. In plain terms: the backbone defines the family, the chains define the individual, and the locants define the position of every feature.

Shorthand codes versus exact names

Lipid shorthand exists because exact names are unreadable at scale. A lipidomics dataset with 8,000 features cannot display full systematic names in every column, so the field converged on string codes. The most important convention is the colon-pair format, total carbons followed by total double bonds. PC 34:1 means a phosphatidylcholine with 34 acyl carbons and one double bond across both chains combined. It could be 16:0/18:1, 18:0/16:1, or 16:1/18:0, and the shorthand alone cannot tell you which.

This is the single largest source of error in lipid reporting. A study claiming that PC 34:1 rises in disease has identified a compositional cluster, not a molecule. Only species-level annotation, with the slash and explicit chain assignments, supports mechanistic claims. Where the analytical method cannot resolve sn-position, the honest notation uses an underscore, as in PC 16:0_18:1, signaling that positional assignment was not determined.

The LIPID MAPS classification and its categories

Since 2005, the LIPID MAPS consortium has organized all lipids into eight categories, and this framework underpins nearly every lipid database and mass-spectrometry library in current use. The categories are fatty acyls, glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, prenol lipids, saccharolipids, and polyketides. Each category carries a two-letter code, so a triacylglycerol is a GL, a phosphatidylcholine is a GP, and a ceramide is an SP.

The category system also fixes how abbreviations get built. The headgroup class comes first (PC, PE, PS, PI, PG, PA for the common glycerophospholipids), then the composition code. The numbering below shows how a single molecule can escalate in specificity from a two-letter class code to a full structural string. Each step adds letters and locants, and each step cuts the number of molecules that could match.

Some codes look similar and are not. PE stands for phosphatidylethanolamine in glycerophospholipid shorthand, while PE also appears in sphingolipid tables as phosphoethanolamine, a different structural context. Ether lipids insert an O- or P- prefix (O- for the plasmanyl alk-1-enyl ether, P- for the plasmalogen vinyl ether). A plasmalogen phosphatidylethanolamine is therefore written PE(P-16:0/18:1) or, in older literature, as a plasmalogen with a vinyl-ether bond at sn-1. Misreading that P- as a phosphate group is a recurring transcription mistake.

Where naming systems diverge: trivial names, isomers, and edge cases

Trivial names remain in legal and clinical use because they are older than the systematic system and often encode history. Arachidonic acid is 20:4(5Z,8Z,11Z,14Z), and the name comes from Arachis, the peanut genus, not from the molecule's chemistry. Linoleic acid is 18:2(9Z,12Z), an omega-6 fatty acid, while alpha-linolenic acid is 18:3(9Z,12Z,15Z), an omega-3. The omega nomenclature counts from the methyl end, the delta nomenclature counts from the carboxyl end, and the two systems assign different numbers to the same double bond. A 20-carbon fatty acid with a double bond at delta-5 is also omega-15, and only one of those numbers appears in most food-label literature.

Isomers create the sharpest edge cases. Oleic acid (18:1, 9Z) and elaidic acid (18:1, 9E) share every atom and locant and differ only in geometry. The trans isomer has a melting point near 113 degrees Fahrenheit versus about 55 degrees Fahrenheit for the cis form, a roughly 58-degree gap that explains why trans fats behave like saturated fats in membranes and in food texture. A nomenclature that omitted the Z/E descriptor would erase a physical distinction with measurable biological consequences.

Branching and unusual linkages compound the problem. Tuberculostearic acid is 10-methyloctadecanoic acid, an iso- or anteiso-branched chain depending on whether the methyl sits at the penultimate or antepenultimate carbon. Cyclopropane fatty acids such as lactobacillic acid (11,12-methylene-octadecanoic acid) carry a ring that changes membrane packing. None of these features survive a sum-composition code, which is why bacterial lipid reports that rely on 18:1 totals lose the diagnostic information entirely.

A short history of lipid nomenclature standards

The systematic framework did not arrive in one step. It accumulated through committee recommendations, database standardization, and analytical necessity, and knowing the sequence helps you read older papers correctly.

The gap between the 1976 recommendations and the 2005 database launch is roughly 29 years, and during that interval most published lipid data used inconsistent or partial naming. A 1995 paper reporting "phosphatidylcholine" without chain detail is not wrong by the standards of its time. It is simply not comparable to a modern species-level dataset, and reanalysis requires care.

Common misconceptions about lipid names

The most persistent misconception is that the number before the colon in a shorthand tells you the carbon count of a single chain. It does not. In a glycerophospholipid code, that first number is the sum across both chains. TAG 52:3 means a triacylglycerol with 52 total acyl carbons, which is consistent with three 16- or 18-carbon chains in various combinations but not with any single chain being 52 carbons long.

A second misconception treats the omega number as a chemical position. The omega-3 and omega-6 families are defined by counting from the methyl terminus, and the designation is a naming convention, not a functional class. Two fatty acids with the same omega number can differ in chain length and in the number of double bonds, and they behave differently. The 2019 recommendation from international lipid organizations to treat omega-3 intake in terms of specific molecules (EPA and DHA) rather than the family label reflects this limitation.

A third misunderstanding confuses the number of double bonds with the number of double-bond positions that must be specified. A shorthand of 20:4 says four double bonds exist but names none of their locations. Without the locant list, 20:4 could describe arachidonic acid, eicosatetraenoic acids of several geometries, or a mixture. Systematic names always carry locants precisely because the shorthand does not.

A fourth, subtler error is treating cis and trans as interchangeable in casual writing. They are not, and the biological difference is large. The trans configuration straightens the chain, raises the melting point, and changes how the lipid packs into a membrane leaflet. Any report that writes 18:1 without geometry has left a material property undefined.

What to actually do when naming or reading lipids

Practical lipid nomenclature comes down to matching the depth of the name to the depth of the claim. If a study measured only total 34:1 composition, the paper should say so. If it resolved chains, it should give the slash notation and, where possible, the sn-assignment. Reviewers and database curators apply exactly this test, and a mismatch between method and name invalidates the comparison.

For anyone building a dataset or writing a methods section, the steps are concrete:

• Report the annotation level explicitly (sum, species, or molecular species) and never upgrade a level in interpretation.
• Use the slash for resolved chains and the underscore for unresolved sn-positions, and state which instrument or standard enabled the assignment.
• Include double-bond geometry (Z/E) whenever the analytical method distinguishes it, and say so if it does not.
• Convert trivial names to systematic names at least once in the text, so readers can map the molecule to a database entry.
• Cross-check every abbreviation against the LIPID MAPS class codes before submission, because PE, PG, and PI collide with unrelated abbreviations in other fields.

For readers, the reverse discipline applies: when a paper reports a shorthand without a slash, treat the finding as compositional. When it reports an exact name with locants, the claim can be evaluated at the molecular level. The name is a contract about what was measured, and honoring that contract is what separates a reproducible lipid result from a suggestive one.

Limits of the system and open questions

The current nomenclature is robust for abundant, well-characterized lipids and weaker at the edges. Oxidized lipids are a clear gap. A hydroperoxide of linoleic acid can form at carbon 9 or carbon 13, and each positional isomer has its own biological activity, yet many databases store oxidized species under generic oxidized-fatty-acid headings that do not distinguish regioisomers. Per the 2009 LIPID MAPS framework, oxidized lipids fall under the fatty acyl category, but the sub-classification for hydroperoxy, hydroxy, and epoxy products remains uneven across databases.

Isotope-labeled lipids and synthetic analogs also strain the conventions. A deuterated internal standard such as d9-PC 34:1 must be named so that the label position is unambiguous, and the standard shorthand has no dedicated slot for isotope placement. Most laboratories append the label in brackets, but this is convention, not a formal rule.

Whether the field will converge on a single machine-readable standard is unresolved. The Lipidomics Standards Initiative has pushed toward harmonized reporting, yet public databases still disagree on shorthand formatting for ether lipids and for branched chains. The hypothesis that a unified identifier system will replace free-text shorthand is plausible but not established, and until it is, the most defensible practice is to report both the shorthand and the systematic name. In plain terms: the code is for computers and tables, the systematic name is for truth, and a careful lipid report needs both.