Annie G. Bryant

Scientists have developed hundreds of techniques to measure the interactions between pairs of processes in complex systems, but these computational methods—from contemporaneous correlation coefficients to causal inference methods—define and formulate interactions differently, using distinct quantitative theories that remain largely disconnected. Here we introduce a large assembled library of 237 statistics of pairwise interactions, and assess their behavior on 1,053 multivariate time series from a wide range of real-world and model-generated systems. Our analysis highlights commonalities between disparate mathematical formulations of interactions, providing a unified picture of a rich interdisciplinary literature. Using three real-world case studies, we then show that simultaneously leveraging diverse methods can uncover those most suitable for addressing a given problem, facilitating interpretable understanding of the quantitative formulation of pairwise dependencies that drive successful performance. Our results and accompanying software enable comprehensive analysis of time-series interactions by drawing on decades of diverse methodological contributions.

Vascular endothelial cells play an important role in maintaining brain health, but their contribution to Alzheimer’s disease (AD) is obscured by limited understanding of the cellular heterogeneity in normal aged brain and in disease. To address this, we performed single nucleus RNAseq on tissue from 32 human AD and non-AD donors (19 female, 13 male) each with five cortical regions: entorhinal cortex, inferior temporal gyrus, prefrontal cortex, visual association cortex and primary visual cortex. Analysis of 51,586 endothelial cells revealed unique gene expression patterns across the five regions in non-AD donors. Alzheimer’s brain endothelial cells were characterized by upregulated protein folding genes and distinct transcriptomic differences in response to amyloid beta plaques and cerebral amyloid angiopathy (CAA). This dataset demonstrates previously unrecognized regional heterogeneity in the endothelial cell transcriptome in both aged non-AD and AD brain.

Background: Reduced cerebrovascular function and accumulation of tau pathology are key components of cognitive decline in Alzheimer’s disease (AD). Recent multimodal neuroimaging studies show a correlation between cortical tau accumulation and reduced cerebral perfusion. However, animal models predict that tau exerts capillary-level changes that may not be fully captured by standard imaging protocols. Objective: Using newly-developed magnetic resonance imaging (MRI) technology to measure capillary- specific perfusion parameters, we examined a series of mild cognitive impairment (MCI) and AD patients with tau positron emission tomography (PET) to observe whole-brain capillary perfusion alterations and their association with tau deposition. Methods: Seven subjects with MCI or AD received Flortaucipir PET to measure tau deposition and spin-echo dynamic susceptibility contrast (SE-DSC) MRI to measure microvascular perfusion (<10μm radius vessels). Gradient-echo (GE) DSC and pseudocontinuous arterial spin labeling (PCASL) MRI were also acquired to assess macrovascular perfusion. Tau PET, microvascular perfusion, and cortical thickness maps were visually inspected in volumetric slices and on cortical surface projections. Results: High tau PET signal was generally observed in the lateral temporal and parietal cortices, with uptake in the occipital cortex in one subject. Global blood flow measured by PCASL was reduced with increasing tau burden, which was consistent with previous studies. Tau accumulation was spatially associated with variable patterns of microvascular cerebral blood flow (CBF) and oxygen extraction fraction (OEF) in the cortex and with increased capillary transit heterogeneity (CTH) in adjacent periventricular white matter, independent of amyloid-β status. Conclusion: Although macrovascular perfusion generally correlated with tau deposition at the whole-cortex level, regional changes in microvascular perfusion were not uniformly associated with either tau pathology or cortical atrophy. This work highlights the heterogeneity of AD-related brain changes and the challenges of implementing therapeutic interventions to improve cerebrovascular function.

Alzheimer’s Disease (AD) is associated with neuropathological changes, including aggregation of tau neurofibrillary tangles (NFTs) and amyloid-beta plaques. Mounting evidence indicates that vascular dysfunction also plays a key role in the pathogenesis and progression of AD, in part through endothelial dysfunction. Based on findings in animal models that tau pathology induces vascular abnormalities and cellular senescence, we hypothesized that tau pathology in the human AD brain leads to vascular senescence. To explore this hypothesis, we isolated intact microvessels from the dorsolateral prefrontal cortex (PFC, BA9) from 16 subjects with advanced Braak stages (Braak V/VI, B3) and 12 control subjects (Braak 0/I/II, B1), and quantified expression of 42 genes associated with senescence, cell adhesion, and various endothelial cell functions. Genes associated with endothelial senescence and leukocyte adhesion, including SERPINE1 (PAI-1), CXCL8 (IL8), CXCL1, CXCL2, ICAM-2, and TIE1, were significantly upregulated in B3 microvessels after adjusting for sex and cerebrovascular pathology. In particular, the senescence-associated secretory phenotype genes SERPINE1 and CXCL8 were upregulated by more than 2-fold in B3 microvessels after adjusting for sex, cerebrovascular pathology, and age at death. Protein quantification data from longitudinal plasma samples for a subset of 13 (n = 9 B3, n = 4 B1) subjects showed no significant differences in plasma senescence or adhesion-associated protein levels, suggesting that these changes were not associated with systemic vascular alterations. Future investigations of senescence biomarkers in both the peripheral and cortical vasculature could further elucidate links between tau pathology and vascular changes in human AD.

Despite a rich and interdisciplinary literature on time-series analysis, the brain’s complex distributed dynamics are typically quantified using only a limited set of manually selected representations and statistics. This leaves open the possibility that alternative dynamical properties may outperform those reported for a given application. Here we address this limitation by introducing a systematic procedure to compare diverse, interpretable features of functional magnetic resonance imaging (fMRI) data, drawn from comprehensive algorithmic libraries to capture both dynamical properties of individual brain regions (intra-regional activity) and patterns of statistical dependence between regions (inter-regional functional coupling). We apply our data-driven approach to investigate alterations to dynamical structures in resting-state fMRI in participants with schizophrenia (SCZ), bipolar 1 disorder (BP), attention-deficit hyperactivity disorder (ADHD), and autism spectrum disorder (ASD). Our findings broadly support the use of linear time-series analysis techniques for fMRI-based case–control analysis, while also identifying novel informative dynamical structures that have previously received less attention. While some simple statistical representations of fMRI dynamics perform surprisingly well (e.g., properties within a single brain region), we found performance improvements through combining intra-regional properties with inter-regional coupling—consistent with distributed, multifaceted changes to fMRI dynamics in neuropsychiatric disorders. The data-driven approach here enables the systematic identification and interpretation of disorder-specific dynamical signatures, with applicability beyond neuroimaging to diverse scientific problems involving complex time-varying systems.