TMT and iTRAQ as concepts

TMT and iTRAQ are trade names of their owners. Laboratories use the names for two isobaric reagent classes that solve the same problem: mix several peptide samples, inject them once, and still read a separate quantity for each sample. This page is a buyer's comparison of the idea, so a specification can be written without pretending that a plex generation or a price is something an article should invent. Bottom-up proteomics in plain language is the identification background. The decision here is whether an isobaric multiplex is the comparison you meant to buy.

One precursor, several reporters

Each tag has two functional parts that matter conceptually. A reporter will become a small ion of a characteristic mass. A balance group makes up the rest of the mass so that every tag in the set has the same total. The same peptide from different samples therefore co-elutes and looks like one precursor. Fragmentation breaks the tag. Reporter ions appear at different masses, often in the low mass region, and their relative heights are the quantities. Backbone fragments of the peptide are what the search engine matches to a sequence.

The instrument has to resolve the reporters the generation requires. Some generations sit so close in mass that only a high-resolution measurement separates them. A specification that names the reagent class and forgets the resolution class will buy a chemistry the mass spectrometer cannot read. That is a purchasing fact, not a biochemical subtlety.

Labelling is usually done on peptides after digestion, on free amines. Incomplete labelling, or labelling in a buffer that still contains amines, puts part of the sample at the wrong mass or at no tag at all. Those peptides drop out of the multiplex or split the signal. The protocol's peptide amount per channel exists because the reagent has a capacity. How laboratories estimate protein concentration matters before that amount is claimed. An estimate blinded by detergent is an uneven multiplex.

What the reporter answers

The reporter answers a relative question: among the channels that were mixed, how did this peptide's signal distribute? It answers that question for the molecules that carried the tag and that fell inside the isolation window. It does not answer how much protein was in the original tissue in absolute units. It does not, by itself, prove that a pathway moved. A protein-level ratio is an inference from the peptides you accepted, under a false discovery rate you should state. Peptides that disagree with each other are a warning, not an average to hide.

A reference channel, often a pool of the samples or a common calibrator mixed into one tag, gives every batch a shared ruler. It is useful when the study is larger than one multiplex. It is also a sample. It consumes protein amount, it must be made the same way each time, and a careless pool dominates the chemistry without representing any one biological group. Write down what the reference is. A diagram of coloured channels copied from a brochure is not that description.

Co-isolation, the central limit

Ratio compression is the limitation that should be in the specification, not in a footnote after the results surprise someone. Peptides that co-elute and whose masses fall inside the isolation window fragment together. Each contributes reporter ions. If one peptide truly differs ten-fold and the neighbour does not differ, the measured reporters move toward a milder ratio. Complex samples, short gradients and wide windows make this worse. Simpler fractions, longer separations and narrower windows make it milder. An extra fragmentation step, on instruments and methods built for it, tries to quantify from fragments of fragments so the neighbour's reporters are left behind. That method class is a real option and a real cost in speed and in sample amount. It is not automatic on every instrument that can "do TMT".

The false discovery rate still applies to the identification. A compressed ratio on a wrongly identified peptide is a confident-looking number on the wrong sequence. Reporter purity does not rescue a bad search, a wrong database, or a modification you forgot to allow. Proteomics Standards Initiative resources are where file formats and reporting expectations for this kind of data are argued in public. They are a standards home. They are not a certificate that a particular multiplex was analysed well.

Carrier channels change the question

A carrier, sometimes called a boost channel, loads one tag with much more peptide than the others, often a mixed biological sample, so that fragment ions are intense enough to identify. The analytical channels may be single cells or other low-input material. Identification improves for a real physical reason: there are more ions of the shared backbone. Quantification of the small channels gets harder for a real physical reason too. Impurities in the carrier tag leak into neighbouring reporters. The carrier's own ratio compression is large. Ion counts in the small channels may be tiny. The experiment can still be the right experiment. It is not "a normal multiplex with one channel bigger". A specification should say whether a carrier is in scope, how the carrier was made, and which claims will be withheld when reporter counts are low.

What the reporter does and does not settle

Trademarks stay with their owners. A quotation can name the reagent class and the generation. It should not be read as a claim that one brand is universally preferable, or that a listed generation is the one currently manufactured. Product generations change. Your sample count does not.

Writing the specification a supplier can answer

State the number of biological samples, the number of replicates, and whether they must fit in one mix or may span several mixes joined by a reference channel. State the protein or peptide amount class per channel, because reagent capacity and mixing ratios both depend on it. State whether a carrier channel is intended. State the instrument resolution class you actually have, because closely spaced reporters are invisible on the wrong analyser. State whether offline fractionation will be used to reduce co-isolation.

Ask the method discussion to name the reagent generation that fits those constraints, the labelling conditions the protocol requires, and how incomplete labelling would be noticed. PRIDE datasets that use these reagent classes show the range of reporting, from careful channel maps to files that never say how the reference was built. Copy the careful habit. UniProt remains the sequence source the search will need; the tag chemistry does not replace it.

Safety and research use

The labelling reagents are chemical hazards with their own safety data sheets. Organic solvent and the hydroxylamine-style quench some protocols use deserve the same respect. Follow the sheet that belongs to the vial you open. An isobaric ratio is a research comparison. It is not a diagnostic result and not a regulatory measurement. Biosafety of the cells or tissues before they were digested is an institutional decision. This comparison does not authorise a clinical claim because several samples were measured "together".

Reagents, heat, and a specification that survives the loading dock

Isobaric reagents are moisture-sensitive and temperature-sensitive in the way the supplier's storage rule describes. A specification written for a warm climate should ask how the reagent is shipped and what you will do if a vial arrives warm or damp. Do not assume a multiplex can be discussed from a product-generation name after the chemistry has already been compromised. Write acceptance checks into the enquiry: appearance as the protocol describes, a labelling-efficiency control if the method includes one, and a rule that an incomplete label stops the study before the instrument queue. ProteomeXchange records show completed studies. They do not show the vials that were discarded on receipt. Your specification should.

What to put on the quote request

Number of samples, need for a reference channel, protein amount per channel, instrument resolution class, whether a carrier is in scope, and whether ratio compression is acceptable for the fold changes you hope to interpret. The differential abundance reference, the shotgun discovery proteomics reference and the protein identification by LC-MS/MS reference are the surrounding method pages. Send the constraints with the quote request and ask for a method discussion. The reply you want names a reagent class and a generation logic. It does not need a slogan about multiplexing.