Stephan Fischer

The cerebral cortex is composed of neuronal types with diverse gene expression that are organized into specialized cortical areas. These areas, each with characteristic cytoarchitecture ^1,2 , connectivity ^3,4 and neuronal activity ^5,6 , are wired into modular networks ^3,4,7 . However, it remains unclear whether these spatial organizations are reflected in neuronal transcriptomic signatures and how such signatures are established in development. Here we used BARseq, a high-throughput in situ sequencing technique, to interrogate the expression of 104 cell-type marker genes in 10.3 million cells, including 4,194,658 cortical neurons over nine mouse forebrain hemispheres, at cellular resolution. De novo clustering of gene expression in single neurons revealed transcriptomic types consistent with previous single-cell RNA sequencing studies ^8,9 . The composition of transcriptomic types is highly predictive of cortical area identity. Moreover, areas with similar compositions of transcriptomic types, which we defined as cortical modules, overlap with areas that are highly connected, suggesting that the same modular organization is reflected in both transcriptomic signatures and connectivity. To explore how the transcriptomic profiles of cortical neurons depend on development, we assessed cell-type distributions after neonatal binocular enucleation. Notably, binocular enucleation caused the shifting of the cell-type compositional profiles of visual areas towards neighbouring cortical areas within the same module, suggesting that peripheral inputs sharpen the distinct transcriptomic identities of areas within cortical modules. Enabled by the high throughput, low cost and reproducibility of BARseq, our study provides a proof of principle for the use of large-scale in situ sequencing to both reveal brain-wide molecular architecture and understand its development.

Genetic and environmental variation are key contributors during organism development, but the influence of minor perturbations or noise is difficult to assess. This study focuses on the stochastic variation in allele-specific expression that persists through cell divisions in the nine-banded armadillo (Dasypus novemcinctus). We investigated the blood transcriptome of five wild monozygotic quadruplets over time to explore the influence of developmental stochasticity on gene expression. We identify an enduring signal of autosomal allelic variability that distinguishes individuals within a quadruplet despite their genetic similarity. This stochastic allelic variation, akin to X-inactivation but broader, provides insight into non-genetic influences on phenotype. The presence of stochastically canalized allelic signatures represents a novel axis for characterizing organismal variability, complementing traditional approaches based on genetic and environmental factors. We also developed a model to explain the inconsistent penetrance associated with these stochastically canalized allelic expressions. By elucidating mechanisms underlying the persistence of allele-specific expression, we enhance understanding of development’s role in shaping organismal diversity.

The cortex is composed of neuronal types with diverse gene expression that are organized into specialized cortical areas. These areas, each with characteristic cytoarchitecture (Brodmann 1909; Vogt and Vogt 1919; Von Bonin 1947), connectivity (Zingg et al. 2014; Harris et al. 2019), and neuronal activity (Schwarz et al. 2008; Ferrarini et al. 2009; He et al. 2009; Meunier et al. 2010; Bertolero et al. 2015), are wired into modular networks (Zingg et al. 2014; Harris et al. 2019; Huang et al. 2020). However, it remains unclear whether cortical areas and their modular organization can be similarly defined by their transcriptomic signatures and how such signatures are established in development. Here we used BARseq, a high-throughput in situ sequencing technique, to interrogate the expression of 104 cell type marker genes in 10.3 million cells, including 4,194,658 cortical neurons over nine mouse forebrain hemispheres at cellular resolution. De novo clustering of gene expression in single neurons revealed transcriptomic types that were consistent with previous single-cell RNAseq studies(Yao et al. 2021a; Yao et al. 2021b). Gene expression and the distribution of fine-grained cell types vary along the contours of cortical areas, and the composition of transcriptomic types are highly predictive of cortical area identity. Moreover, areas with similar compositions of transcriptomic types, which we defined as cortical modules, overlap with areas that are highly connected, suggesting that the same modular organization is reflected in both transcriptomic signatures and connectivity. To explore how the transcriptomic profiles of cortical neurons depend on development, we compared the cell type distributions after neonatal binocular enucleation. Strikingly, binocular enucleation caused the cell type compositional profiles of visual areas to shift towards neighboring areas within the same cortical module, suggesting that peripheral inputs sharpen the distinct transcriptomic identities of areas within cortical modules. Enabled by the high-throughput, low-cost, and reproducibility of BARseq, our study provides a proof-of-principle for using large-scale in situ sequencing to reveal brain-wide molecular architecture and to understand its development.

Interactions between proteins help us understand how genes are functionally related and how they contribute to phenotypes. Experiments provide imperfect ‘ground truth’ information about a small subset of potential interactions in a specific biological context, which can then be extended to the whole genome across different contexts, such as conditions, tissues or species, through machine learning methods. However, evaluating the performance of these methods remains a critical challenge. Here, we propose to evaluate the generalizability of gene characterizations through the shape of performance curves. We identify Functional Equivalence Classes (FECs), subsets of annotated and unannotated genes that jointly drive performance, by assessing the presence of straight lines in ROC curves built from gene-centric prediction tasks, such as function or interaction predictions. FECs are widespread across data types and methods, they can be used to evaluate the extent and context-specificity of functional annotations in a data-driven manner. For example, FECs suggest that B cell markers can be decomposed into shared primary markers (10–50 genes), and tissue-specific secondary markers (100–500 genes). In addition, FECs suggest the existence of functional modules that span a wide range of the genome, with marker sets spanning at most 5% of the genome and data-driven extensions of Gene Ontology sets spanning up to 40% of the genome. Simple to assess visually and statistically, the identification of FECs in performance curves paves the way for novel functional characterization and increased robustness in the definition of functional gene sets. Code for analyses and figures is available at https://github.com/yexilein/pyroc. Supplementary data are available at Bioinformatics online.

Pre-processing of single-cell RNA sequencing datasets is a crucial step for the success of downstream analyses, including visualization, cell type clustering or pseudotime analyses. However, there is currently no consensus on how to account for the variability in the number of RNAs captured per cell. Typically, this potential confounder is handled by applying simple cell-level correction procedures (such as dividing by the total number of counts). Recent alternatives include parametric methods that leverage the shape of gene distributions to provide a normalization that is both cell- and gene-specific (such as sctransform, popularized by the Seurat package). In this study, the authors highlight the risk of overspecifying such models (defining more parameters than can be inferred from the data), which results in biased parameter estimations and lower downstream performance. By combining analytical derivations and data-driven controls, the study elegantly demonstrates how overspecification can be identified and remedied, reducing popular parametric methods to a more parsimonious and easily computable form.

An essential step toward understanding brain function is to establish a structural framework with cellular resolution on which multi-scale datasets spanning molecules, cells, circuits and systems can be integrated and interpreted¹. Here, as part of the collaborative Brain Initiative Cell Census Network (BICCN), we derive a comprehensive cell type-based anatomical description of one exemplar brain structure, the mouse primary motor cortex, upper limb area (MOp-ul). Using genetic and viral labelling, barcoded anatomy resolved by sequencing, single-neuron reconstruction, whole-brain imaging and cloud-based neuroinformatics tools, we delineated the MOp-ul in 3D and refined its sublaminar organization. We defined around two dozen projection neuron types in the MOp-ul and derived an input–output wiring diagram, which will facilitate future analyses of motor control circuitry across molecular, cellular and system levels. This work provides a roadmap towards a comprehensive cellular-resolution description of mammalian brain architecture.

Machine learning in genomics plays a key role in leveraging high-throughput data, but assessing the generalizability of performance has been a persistent challenge. Here, we propose to evaluate the generalizability of gene characterizations through the shape of performance curves. We identify Functional Equivalence Classes (FECs), uniform subsets of annotated and unannotated genes that jointly drive performance, by assessing the presence of straight lines in ROC curves. FECs are widespread across modalities and methods, and can be used to evaluate the extent and context-specificity of functional annotations in a data-driven manner. For example, FECs suggest that B cell markers can be decomposed into shared primary markers (10 to 50 genes), and tissue-specific secondary markers (100 to 500□genes). In addition, FECs are compatible with a wide range of functional encodings, with marker sets spanning at most 5% of the genome and data-driven extensions of Gene Ontology sets spanning up to 40% of the genome. Simple to assess visually and statistically, the identification of FECs in performance curves paves the way for novel functional characterization and increased robustness in analysis.

High-throughput, spatially resolved gene expression techniques are poised to be transformative across biology by overcoming a central limitation in single-cell biology: the lack of information on relationships that organize the cells into the functional groupings characteristic of tissues in complex multicellular organisms. Spatial expression is particularly interesting in the mammalian brain, which has a highly defined structure, strong spatial constraint in its organization, and detailed multimodal phenotypes for cells and ensembles of cells that can be linked to mesoscale properties such as projection patterns, and from there, to circuits generating behavior. However, as with any type of expression data, cross-dataset benchmarking of spatial data is a crucial first step. Here, we assess the replicability, with reference to canonical brain subdivisions, between the Allen Institute’s in situ hybridization data from the adult mouse brain (Allen Brain Atlas (ABA)) and a similar dataset collected using spatial transcriptomics (ST). With the advent of tractable spatial techniques, for the first time, we are able to benchmark the Allen Institute’s whole-brain, whole-transcriptome spatial expression dataset with a second independent dataset that similarly spans the whole brain and transcriptome. We use regularized linear regression (LASSO), linear regression, and correlation-based feature selection in a supervised learning framework to classify expression samples relative to their assayed location. We show that Allen Reference Atlas labels are classifiable using transcription in both data sets, but that performance is higher in the ABA than in ST. Furthermore, models trained in one dataset and tested in the opposite dataset do not reproduce classification performance bidirectionally. While an identifying expression profile can be found for a given brain area, it does not generalize to the opposite dataset. In general, we found that canonical brain area labels are classifiable in gene expression space within dataset and that our observed performance is not merely reflecting physical distance in the brain. However, we also show that cross-platform classification is not robust. Emerging spatial datasets from the mouse brain will allow further characterization of cross-dataset replicability ultimately providing a valuable reference set for understanding the cell biology of the brain.

Summary: Our understanding of cell types has advanced considerably with the publication of single cell atlases. Marker genes play an essential role for experimental validation and computational analyses such as physiological characterization through pathway enrichment, annotation, and deconvolution. However, a framework for quantifying marker replicability and picking replicable markers is currently lacking. Here, using high quality data from the Brain Initiative Cell Census Network (BICCN), we systematically investigate marker replicability for 85 neuronal cell types. We show that, due to dataset-specific noise, we need to combine 5 datasets to obtain robust differentially expressed (DE) genes, particularly for rare populations and lowly expressed genes. We estimate that 10 to 200 meta-analytic markers provide optimal performance in downstream computational tasks. Replicable marker lists condense single cell atlases into interpretable and generalizable information about cell types, opening avenues for downstream applications, including cell type annotation, selection of gene panels and bulk data deconvolution.