Gaussian process models underlie many spatial transcriptomics tools but typically assume stationary covariance. To test this assumption, we use approximate Bayes factors from R-INLA to compare stationary and non-stationary Matérn covariance functions across 12 Visium datasets; covariance non-stationarity is substantially evident as it appears in 3% to 50% of genes across datasets. We find that gene sets associated with effector functions are enriched among genes favoring non-stationary spatial covariance.

Background: Down syndrome (DS) is associated with persistent language and cognitive impairments and with abnormalities in cortical oscillatory activity, including in the gamma range. Transcranial photobiomodulation (tPBM) is a noninvasive neuromodulatory intervention with potential benefits for cortical physiology, language, and cognition. Methods: We conducted a pilot randomized, double-blind, sham-controlled trial of repeated 40-Hz near-infrared tPBM in adolescents and young adults with DS. Fourteen participants were randomized 1:1 to active tPBM or sham and received 18 sessions over 6 weeks, followed by short-term and long-term follow-up. Outcomes included resting-state EEG gamma power, connected-speech measures, language and cognitive indices, and selected computerized tasks. Results: Active tPBM did not significantly increase global resting-state EEG gamma power relative to sham at either follow-up. Pre-registered analyses did not show broad treatment benefit across outcomes, although they did identify a significant short-term advantage for active tPBM on grammatical morpheme accuracy in connected speech; in contrast, picture naming favored sham at long-term follow-up. Exploratory mechanistic analyses did not show a robust biological treatment signal. Both active and sham procedures were well tolerated, with no serious adverse events. Conclusions: In this underpowered pilot sample, 6 weeks of 40-Hz near-infrared tPBM–delivered unilaterally, at low power, on target areas–did not demonstrate a meaningful effect size in DS. A dose-finding study for tPBM in DS, also accounting for age of participants, is recommended.

Neural activity is organized in low-dimensional structure, yet functional connectivity in fMRI typically represents each brain parcel by a single voxel-averaged time series. This scalar representation makes whole-brain connectivity tractable but discards potentially informative dimensions of within-parcel BOLD activity. Here, we represent each parcel by a low-dimensional temporal subspace derived from its principal-component time series and use the RV coefficient to quantify connectivity between regional subspaces. Across Human Connectome Project resting-state and working-memory data, progressively expanding these subspaces reveals reproducible connectivity regimes with distinct network and identifiability profiles. At rest, connectivity constructed from the first principal component identifies individuals more strongly than either voxel-averaged functional connectivity or higher-dimensional subspace representations. During working memory, identifiability instead peaks after secondary components are included, indicating that the distribution of individual-specific information across the regional PCA spectrum depends on cognitive state. These patterns replicate across independent samples and remain robust across multiple parcellation resolutions. Together, our findings show that within-parcel BOLD structure contains identity- and state-dependent information that is obscured by scalar regional summaries. Functional connectivity may therefore be better understood as a family of related connectomes indexed by the regional subspace retained, providing a general framework for mapping interactions between low-dimensional neural representations.