Coming back to: Why quantitation is the hard part, not detection

Why quantitation is the hard part, not detection — that is the question, and I have not found it answered plainly anywhere I have looked.

Posting the method first, because I know what the first three replies will otherwise be.

• Column: C18, 2.1 x 100 mm, 1.7 um
• Mobile phase: 0.1% TFA in water / 0.1% TFA in acetonitrile
• Gradient: 12% to 45% organic over 22 minutes
• Detection: 214 nm
• Injection: 12 uL
• Sample: retatrutide, reconstituted to 0.5 mg/mL, injected within an hour

The main peak integrates at 98.2% of total area. There is a small feature on the trailing edge that I cannot decide is a shoulder or a baseline artefact, and that is what I am actually asking about.

On the opening post — agreed on the reasoning, with one qualification.

Lateral-flow devices: like a rapid COVID test. They have a reagent strip and produce a colour result. They are quick but not precise and not intended for quantitative work.

Sample handling matters: if a sample is contaminated or degraded before testing, the test gives you the result for the degraded sample, not for fresh material. That is not a failure of the test; it is the test working correctly.

False positives and false negatives: no test has 100% sensitivity and 100% specificity. A negative result does not guarantee absence; a positive result does not guarantee presence. The predictive value depends on prevalence.

Molecular tests: DNA or RNA tests are not applicable to peptides, which do not contain nucleic acids. Peptide-specific tests rely on protein recognition, not nucleic-acid detection.

Worth separating two things that post #2 runs together.

Enzyme assays: if you have an enzyme that reacts specifically with your compound, an enzyme-based assay can work. These are sometimes available commercially but are less common than antibody-based tests.

This follows post #4 rather than contradicting it.

What physical quantity does the device measure? A colorimetric or lateral-flow method typically responds to a class of chemistry rather than to a specific molecule. It can distinguish "peptide present" from "nothing present" but generally cannot distinguish the intended peptide from a closely related one.

Why third-party testing is stronger: a home test run by the person who made the compound is a self-selected result. The test run by a neutral third party removes obvious sources of bias.

post #8 answers the question as asked. The question underneath it is different.

Immunoassay limitations: tests that use antibodies have cross-reactivity limitations. An antibody raised to semaglutide will cross-react to some degree with tirzepatide and other structurally similar compounds. The test result conflates them.

Precision of home tests: most consumer-accessible devices are not precise to the level people want ("is this 99% pure or 97% pure"). They give a yes/no or a rough category, not a quantitative number.

Colorimetric tests: colour intensity correlates with concentration, but the relationship is non-linear at extremes and the correlation depends on the exact conditions (pH, temperature, timing). Reading the colour visually introduces subjectivity.

What you can use home tests for: confirmation that a solution contains peptide (as opposed to being saline alone), rough screening for very different compounds (semaglutide versus totally unrelated compound). What you cannot use them for: precise purity assessment or verification that the identity is correct.

On post #9 — agreed on the reasoning, with one qualification.

What you can use home tests for: confirmation that a solution contains peptide (as opposed to being saline alone), rough screening for very different compounds (semaglutide versus totally unrelated compound). What you cannot use them for: precise purity assessment or verification that the identity is correct.

False positives and false negatives: no test has 100% sensitivity and 100% specificity. A negative result does not guarantee absence; a positive result does not guarantee presence. The predictive value depends on prevalence.