02.01 / Scientific foundations
Individual analysis is the first step in CCRM: available person-specific findings are checked for relevance, quality and timing, then organized into a structured profile.
The workflow starts by reviewing available individual clinical, molecular, laboratory, medication and time-related data. Only relevant available findings enter the profile; no fixed test panel is implied. Their source, collection time, units and uncertainty are retained. A blood measurement is not a direct measurement of a tumor or of every cell.
Map equivalent concepts to shared identifiers, convert compatible units and preserve original values. Record assay methods, reference ranges and tissue origin. Batch effects and differences between laboratories can imitate biological differences; harmonization must be checked rather than assumed. [1]
An unavailable result must remain unknown, not become zero or a normal value. Imputation, when justified, requires documented assumptions and sensitivity checks. For predictive studies, preprocessing must be learned on training data only to avoid information leakage into evaluation. [3]
Align observations to treatment cycles and sampling times. Medication and supplement exposure, liver and kidney function, and changes over time may alter interpretation. Supplements can interact with cancer drugs through metabolism or transport; any real application requires clinical review. [2]
Toggle metadata fields for a fictional measurement. A missing unit or time cannot be repaired by a larger amount of unrelated data.
The demonstration shows metadata completeness only. It does not measure health, cancer risk or suitability for a compound.
References support the methods described, not the validation or performance of Curetanics. Source titles are shown in their original language.