CORE clonotypes: a reproducible structural component of the peripheral TCR repertoire revealed through variation in blood processing times

High-throughput T-cell receptor (TCR) repertoire sequencing is widely used for immune profiling, yet repertoire measurements remain sensitive to pre-analytical variables that can affect reproducibility and downstream interpretation. Among these variables, delays between blood collection and sample processing may alter detectable repertoire composition, particularly among low-frequency clonotypes. Here, we use processing-delay variability as a structured technical perturbation to examine organizational features of the peripheral TCR repertoire. We analyzed peripheral blood TCRβ repertoires from 211 patients with newly diagnosed colorectal cancer, comparing samples processed shortly after blood collection with samples subjected to extended processing delays. Delayed processing was associated with reduced clonotype richness, depletion of singleton clonotypes, and altered diversity distributions, indicating preferential loss of low-frequency repertoire elements under prolonged handling conditions. Within this framework, we defined CORE clonotypes (COnsistent Repertoire Elements) operationally as amino-acid-identical CDR3β sequences detected in both processing groups. We characterized CORE clonotypes across multiple analytical levels, including abundance distributions, V/J gene usage, amino-acid composition, nucleotide redundancy, codon usage, biochemical properties, and correlation-network structure. CORE clonotypes were largely distinct from abundance-matched clonotype sets, indicating that reproducible detection across perturbation conditions is not explained solely by clonotype abundance. At the nucleotide level, CORE clonotypes exhibited elevated synonymous sequence diversity while maintaining conserved amino-acid composition across processing conditions. Following sequencing-depth normalization, CORE clonotypes showed reduced codon-usage divergence and narrower distributions across multiple biochemical properties relative to group-unique clonotypes. Correlation-network analysis further identified a small CORE-associated subnetwork enriched for previously annotated antigen associations compared with randomized controls. These findings suggest that technical perturbations can reveal reproducibly detected, non-random repertoire elements with distinct organizational properties across multiple molecular levels. More broadly, this work highlights the value of pre-analytical variability as a tool for characterizing reproducible repertoire structure in immune repertoire sequencing studies.