Unit Harmonization Before Peptide Assay Data Review

This research article is published on September 3, 2026. It addresses administrative data preparation for peptide assay review. It does not validate an assay, approve a conversion, judge a result, or replace scientific quality review.

The question behind the data table

When peptide assay results arrive from more than one laboratory, how can a team quantify unit and metadata conflicts before a scientist compares the values? A column headed "concentration" can contain values reported as ng/mL, nmol/L, percent label claim, or a laboratory-specific calculated output. Even two identical unit strings may conceal different analytes, matrices, calibration models, or molecular assumptions.

The risk is not limited to arithmetic. A tidy normalized table can create false confidence if it drops the original unit or loses the source record. The research question is therefore about a controlled intake process: what can administrative review detect without making scientific transformations?

Distinguishing three types of traceability

NIST describes metrological traceability as a property of a measurement result linked to a reference through a documented chain, with each calibration contributing to uncertainty. NIST also distinguishes that concept from chain of custody, which tracks possession or handling. FDA materials provide context for data integrity and bioanalytical method validation.

For an intake study, three trails should remain separate. Source traceability connects the row to the submitted report or data file. Transformation traceability records any approved mapping or conversion applied after receipt. Metrological traceability concerns the measurement result and its reference chain. A complete source link does not prove metrological traceability, and a calibrated instrument name does not settle whether a later spreadsheet conversion was correct.

Method: preserve first, classify second

Create an immutable intake layer before attempting harmonization. Retain the submitted value, literal submitted unit, analyte label, matrix, sample identifier, method identifier, laboratory, report version, and source location. Hashes or controlled document identifiers can strengthen the link where the client's system permits them. Never overwrite the original field with a converted number.

Next, compare fields against a scientist-approved data dictionary. The dictionary should define accepted labels and permitted synonyms, not merely a list of preferred spellings. A mapping between "micrograms per milliliter" and "µg/mL" may be lexical. A conversion between mass and molar concentration requires scientific inputs such as the applicable molecular mass and analyte definition. That second case must stay outside an administrator's discretion.

Classify each row with one of these evidence states:

1. Exact dictionary match, where the submitted unit and context agree with an approved entry.
2. Approved label synonym, where only the written form differs and the mapping is documented.
3. Conversion required, where a qualified scientist must approve the rule and inputs.
4. Context conflict, where analyte, matrix, method, or report metadata disagree.
5. Unresolved, where the source lacks enough information to classify the record.

Do not treat the first two states as proof that results are comparable. They only show that administrative metadata passed the defined intake rule.

What to measure

Measure conflict prevalence with the number of eligible rows and files visible. File-level and row-level rates answer different questions. One malformed file can generate thousands of row conflicts, while ten small files can each require separate follow-up. Report both when workload planning depends on them.

Useful measures include first-pass dictionary match, files with at least one unresolved unit, rows awaiting scientific conversion approval, median age of open questions, and records reopened after a mapping change. Count the number of unique unapproved labels because repeated exceptions may indicate a missing dictionary entry rather than repeated independent mistakes.

Time studies should distinguish mechanical parsing, administrative investigation, scientific review, and dependency waiting. A fast conversion is not evidence of a safe conversion. If a source correction arrives, preserve the earlier report version and record why the active record changed.

A practical exception register

An exception register should be compact enough to audit. Each line needs a restricted record reference, received label, expected dictionary context, discrepancy class, owner, opened date, evidence link, disposition, approver where required, and closure date. Free-text notes should state what conflicts, not speculate about the scientific cause.

For example, "unit listed as ng/mL in CSV and nmol/L in signed report" is a factual discrepancy. "CSV is wrong" is a conclusion that the evidence does not yet support. The first wording keeps the issue open for the correct owner.

PeptideStaff can prepare that register, check whether required fields are present, route exceptions, and produce aging summaries. PeptideStaff should not calculate a molecular conversion, select a molecular weight, determine which result is scientifically correct, approve a calibration chain, or release a data set for analysis.

Evidence scope and limitations

The cited sources explain traceability, data integrity, and bioanalytical validation principles. They do not establish a standard conflict rate or administrative staffing ratio for peptide assay intake. A local study can measure its own mix of laboratories and formats, but the result should not be generalized beyond that setting.

Automated parsers can mishandle superscripts, Greek characters, locale-specific decimal marks, or units embedded in notes. A dictionary may be internally consistent and still scientifically incomplete. Sampling only finalized reports will miss the effort spent correcting drafts. Counting every repeated row can exaggerate independent workload. These limitations belong beside the results.

Conclusion from the evidence

Unit harmonization should begin with preserved source values and end with a documented, authorized transformation trail. Administrative measurement is useful because it shows where labels and context fail to line up. It cannot establish scientific comparability. PeptideStaff's defensible role is to organize literal evidence and unresolved questions so qualified laboratory and scientific owners can make the decisions.