Orthogonal Decomposition of Multi-Tissue Aging Clocks Disentangles Deterministic Gene Regulation, Intracellular Stochastic Drift, and Extrinsic Compositional Remodeling

Transcriptomic and epigenetic aging clocks predict chronological and biological age with high precision, yet conventional models collapse cell-extrinsic tissue remodeling, cell-intrinsic regulated stress responses, and intracellular stochastic noise into a single scalar score. This study introduces OrthoClock (orthoclock), an open-source statistical framework that decomposes multi-tissue aging signatures into three orthogonal axes: Extrinsic Compositional Remodeling (Delta_extrinsic) in centered log-ratio simplicial space, Intrinsic Deterministic Regulation (Delta_deterministic) within purified cell types, and Intrinsic Orthogonal Stochastic Drift (Delta_stochastic) orthogonalized (r = 0.0000) against signed and magnitude mean shifts via bivariate Gram-Schmidt projection. In generative Negative-Binomial transcriptional bursting simulations (50 Monte Carlo replicates x 2,000 genes), bivariate Gram-Schmidt projection, strict in-fold young-centroid cross-validation, and independent split-sample perturbation estimation eliminate severe false-positive inflation caused by bursty mean-variance coupling (35.3% down to 6.1%), pre-cross-validation feature leakage (r = 0.661 down to -0.014 under a pure null), and shared-control regression-to-the-mean (r = -0.501 down to -0.002 under a null intervention). Across 308,539 single cells, 947 lifespan bulk RNA-seq profiles, 14 human epigenetic clocks (4,800 CpGs), and 1,128 human methylomes, leak-free in-fold stochastic noise independently predicts single-cell chronological age (r = 0.532), improves combined OrthoClock accuracy in 95.3% of cell types (r = 0.771), and selectively targets pre-mRNA splicing rather than ribosomal quality control. In human blood methylomes, enhancer clock CpGs act as extrinsic leukocyte sensors (p = 1.20e-6), whereas evolutionarily conserved Polycomb (PRC2) clock CpGs (p = 0.0384) exhibit 12.6-fold higher stochastic variance expansion (p = 6.24e-11) linked to pan-tissue downregulation of chromatin maintenance factors (ATRX, SUZ12, TET1/2/3, SIRT1). In split-sample heterochronic parabiosis, young blood reverses deterministic aging across 20 of 20 organs (75.1% of genes reversed) by quieting a vascular-macrophage paracrine pacemaker, but reverses orthogonal stochastic drift significantly less completely across organs (66.2%, p = 0.0020) and 70 purified cell-lineage strata (57.9% vs. 66.4%, p = 2.42e-6). Complete rejuvenation therefore requires pairing systemic anti-inflammatory modulation with cell-intrinsic chromatin maintenance restoration.