Mardiyyah/TAPT_data_V2_split · Datasets at Hugging Face

PMCID stringclasses 24 values	Title stringclasses 24 values	Sentences stringlengths 2 40.7k
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	For projecting to HNOCA, the query model was based on the scPoli model pretrained with the HNOCA data, and fineturned with a batch size of 16,384 for a maximum of 30 epochs with 20 pretraining epochs.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	A nearest neighbour graph was created for the disease-modelling atlas on the basis of the projected latent representation to HNOCA with scanpy (default parameters), with which a UMAP embedding was created with scanpy (default parameters).
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Next, for both HNOCA and the disease-modelling atlas, cells were represented by the concatenated representation of HNOCA-scPoli and primary-scANVI models.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	A bipartite wkNN graph was then reconstructed as mentioned above, by identifying 50 nearest neighbours in HNOCA for each disease-modelling atlas cell.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	On the basis of the bipartite wkNN, the majority voting-based label transfer was applied to transfer the four levels of hierarchical cell type annotation and regional identity to the disease-modelling atlas.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	For each cell in the disease-modelling atlas, a matched HNOCA metacell was reconstructed on the basis of the above mentioned bipartite wkNN.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	In brief, for a query cell i and a gene j measured in HNOCA, its matched metacell expression of j, denoted as , is calculated as:[12pt] $$_}^ }=__}_}_}}__}_}}$$′=∑k⊆Niwikekj∑k⊆NiwikHere, Ni represents all HNOCA nearest neighbours of the query cell ci, wik represents the edge weight between query cell i and reference cell k, and ekj represents expression level of gene j in reference cell k. Given the matched HNOCA metacell transcriptomic profile, the similarity between a query cell and its matched cell state in HNOCA is then calculated as the Spearman correlation between the query cell transcriptomic profile and its matched HNOCA metacell transcriptomic profile.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	To analyse the glioblastoma organoid dataset (GBM-2019), cells from the publication were subset from the integrated disease-modelling atlas.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Using scanpy, highly variable genes were identified with default parameters.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	The log-normalized expression values of the highly variable genes were then scaled across cells, the truncated PCA was performed with the top 20 principal components used for the following analysis.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Next, harmonypy, the Python implementation of harmony, was applied to integrate cells from different samples.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	On the basis of the harmony-integrated embeddings, the neighbour graph was reconstructed.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	UMAP embeddings and Louvain clusters (resolution, 0.5) were created on the basis of the nearest neighbour graph.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Among the 12 identified clusters, cluster-7 and cluster-0, the two clusters with the highest AQP4 expression, were selected for the following DE analysis.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	To analyse the FXS dataset (FXS-2021), cells from the publication were subset from the integrated disease-modelling atlas.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	The same procedure of highly variable gene identification, data scaling and PCA as the GBM-2019 dataset was applied.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Next, the nearest neighbour graph was created directly on the basis of the top 20 principal components.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	UMAP embeddings and Louvain clusters (resolution, 1) were then created on the basis of the reconstructed nearest neighbour graph.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Among the 30 clusters, cluster-17 and cluster-23, which express EMX1 and FOXG1 and were largely predicted to be dorsal telencephalic NPCs and neurons according to the transferred labels from HNOCA, were selected for the following DE analysis.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	To compare expression levels of two groups of paired cells, the expression difference per gene of each cell pair is first calculated on the basis of the log-normalized expression values.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Next, for each gene to test for DE, its variance over the calculated expression difference per cell pair (σ) is compared with the sum of squared of expression differences (di for gene i) normalized by the number of cell pairs:[12pt] $$_^=_^_}.$$02=∑i=1ndin.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Here, an F-test is applied for the comparison, with f = σ/s0, d.f.1 = n − 1 and d.f.2 = n. To construct the HNOCA-CE, we first collected raw count matrices and associated metadata of five more neural organoid studies.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	For two publications, we obtained them from the sources listed in the ‘Data availability’ section of the paper.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	For the remaining three publications, count matrices and associated metadata were provided directly by the authors.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	We subset each dataset to the healthy control cells and removed any cells with fewer than 200 genes expressed.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	We subset the gene space of every dataset to the 3,000 HVGs of HNOCA while filling the expression of missing genes in the community datasets with zeros.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	On average, 23% of genes with zero expression were added per dataset.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	We instantiated a mapping object from the HNOCA-tools package (at commit fe38c52) using the saved scPoli model weights from the HNOCA integration.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Using the map_query method of the mapper instance, we projected the community datasets to HNOCA.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	We used the following training hyperparameters: retrain = ‘partial’, batch_size = 256, unlabeled_prototype_training = False, n_epochs = 10, pretraining_epochs = 9, early_stopping_kwargs = early_stopping_kwargs, eta = 10, alpha_epoch_anneal = 10.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	We computed the wkNN graph using the compute_wknn method of the mapper instance with k = 100.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	We transferred the final level_2 cell type labels from HNOCA to the community datasets using this neighbour graph.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	To obtain the combined representation of HNOCA-CE, we projected HNOCA together with the added community datasets through the trained model and computed a neighbour graph and UMAP from the resulting latent representation.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
PMC11578878	An integrated transcriptomic cell atlas of human neural organoids.	Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at 10.1038/s41586-024-08172-8.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Deciphering patterns of connectivity between neurons in the brain is a critical step toward understanding brain function.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Imaging-based neuroanatomical tracing identifies area-to-area or sparse neuron-to-neuron connectivity patterns, but with limited throughput.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Barcode-based connectomics maps large numbers of single-neuron projections, but remains a challenge for jointly analyzing single-cell transcriptomics.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Here, we established a rAAV2-retro barcode-based multiplexed tracing method that simultaneously characterizes the projectome and transcriptome at the single neuron level.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	We uncovered dedicated and collateral projection patterns of ventromedial prefrontal cortex (vmPFC) neurons to five downstream targets and found that projection-defined vmPFC neurons are molecularly heterogeneous.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	We identified transcriptional signatures of projection-specific vmPFC neurons, and verified Pou3f1 as a marker gene enriched in neurons projecting to the lateral hypothalamus, denoting a distinct subset with collateral projections to both dorsomedial striatum and lateral hypothalamus.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	In summary, we have developed a new multiplexed technique whose paired connectome and gene expression data can help reveal organizational principles that form neural circuits and process information.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Wiring diagrams of a brain can be divided into three levels: (1) the macroscale connectome that describes inter-areal connections, (2) the mesoscale connectome that describes connections between cells, and (3) the microscale connectome that describes connections at the synaptic level (Zeng, 2018).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Studying circuit architecture at the level of the mesoscale connectome describes how information flows between brain regions (Oh et al., 2014).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Traditionally, neuroanatomical tracers are used to characterize regional connectivity matrices (Cowan, 1998).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	To obtain cell-type-specific connectivity, one can use recombinant virus-based tracer in transgenic model organisms or more precisely trace a specific component of a neural circuit using viral-genetic tracing tools to dissect the input-output organization (Ghosh et al., 2011; Nassi et al., 2015; Schwarz et al., 2015).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	However, these methods are highly reliant on complex recombinant virus design and genetically modified model organism, and often are not at a single-neuron resolution.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	While recent advances have brought invaluable insights into understanding neuronal circuits at single-neuron resolution, existing methods have limitations.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	High-throughput fluorescence imaging, such as fluorescence micro-optical sectioning tomography (fMOST), can reconstruct detailed neuron morphologies but requires specialized expertise and equipment and lack transcriptomic information (Gong et al., 2016; Rompani et al., 2017).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Barcode-based methods like MAPseq, BRICseq (multiplexed MAPseq), BARseq, and ConnectID utilize sequencing to map projections (Chen et al., 2019; Huang et al., 2020; Kebschull et al., 2016; Klingler et al., 2021).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	However, MAPseq and BRICseq can only provide connectome information (Huang et al., 2020; Kebschull et al., 2016), BARseq is constrained to assessing a handful of genes via in situ hybridization (Chen et al., 2019), and ConnectID has low recovery of cells with dual connectome-transcriptome data (~16%, 391 cells with connectome barcode identity in 2450 cells with scRNA-seq; Klingler et al., 2021).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	VECTORseq, a Retro-seq-based method (Tasic et al., 2018), is limited by its number of transgenic barcodes used (Cheung et al., 2021).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	The updated BARseq protocol enables detection of up to 100 genes, but throughput remains lower and oligo synthesis costs remain higher compared to scRNA-seq (Chen et al., 2023; Sun et al., 2021).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	In summary, despite significant progress, existing methods fall short in efficiently integrating high-throughput projectomes and transcriptomes at the single-neuron level, hindering a comprehensive understanding of the connectomic and transcriptomic interplay in neuronal circuitry.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Medial prefrontal cortex (mPFC) is an intricate brain region involved in higher order cognitive functions, information processing (e.g., memory and emotions) and driving goal-directed actions (Le Merre et al., 2021).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	For example, mPFC neurons projecting to the nucleus accumbens encoding punishment-related internal states were located in more superficial layer 5a, and mPFC neurons projecting to the ventral tegmental area encoding aversive learning were located in deeper layer 5b (Kim et al., 2017; Wu et al., 2021).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Although previous studies have extensively investigated the anatomical and functional diversities of mPFC, the relationship between anatomical and molecular features of mPFC neurons remains elusive.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Do mPFC neurons projecting to different downstream brain regions differ in their transcriptomes?
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Are these projection-defined mPFC neurons homogeneous or composed of different neuron subtypes?
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	The answer to these questions may be further complicated by the finding that mPFC neurons can send collateral axons to multiple brain regions (Cornwall and Phillipson, 1988).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	So, what are the principles of target selection or target combination for these collateral projection mPFC neurons?
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	What are the cell type and molecular features of these ‘broadcasting’ neurons?
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	To address these challenges, we designed a multiplexed tracing method capable of characterizing single-neuron transcriptome and projectome at the same time, which we called MERGE-seq (Multiplexed projEction neuRons retroGrade barcodE).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	We used MERGE-seq to interrogate the projectome and the corresponding transcriptome of ventral mPFC neurons.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	We injected five rAAV2-retro viruses with distinct barcodes into the five known downstream targets of ventromedial prefrontal cortex (vmPFC), including agranular insular cortex (AI), dorsomedial striatum (DMS), basolateral amygdala (BLA), mediodorsal thalamic nucleus (MD), and lateral hypothalamus (LH), in the same mouse brain such that each target region received a unique barcoded rAAV2-retro.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	We found that vmPFC neurons projecting to each downstream target are heterogeneous, which are composed of transcriptionally different subtypes of neurons.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Approximately 65% of barcoded vmPFC neurons exhibited dedicated projection patterns based on MERGE-seq data, sending axonal projections exclusively to one of the five selected targets.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	It is important to note that this characterization of ‘dedicated projection’ neurons is specifically defined in the context of the five target regions examined in this study.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Approximately 35% of barcoded vmPFC neurons sent collateral projections to multiple brain regions, most of which are dual-target projection neurons (bifurcated projection).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	We further uncovered the cell type compositions and layer distributions of these dedicated and collateral projection vmPFC neurons, and revealed their molecular signatures.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	We validated complex MERGE-seq-inferred projection patterns by joint analysis with recently published single-neuron projectome data (Gao et al., 2022).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Additionally, dual-modal interrogation using RNA fluorescence in situ hybridization (FISH) and dual-color retrograde AAV labeling allowed us to confirm vmPFC neuron bifurcations to DMS and LH, demonstrating layer 5 Pou3f1 neurons collateralize between these targets.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Finally, we implemented a machine learning-based methodology and uncovered specific gene clusters for predicting certain projection patterns.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	As MERGE-seq bridges the gap between single-neuron projectome and transcriptome data, it can uncover new molecular properties of anatomical neural circuits.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	In order to use the 10x Genomics scRNA-seq system to analyze transcripts from cells infected with rAAV2-retro virus, we modified the viral vector by adding a 15 bp barcode index and polyadenylation signal sequences to the 3’ end of the EGFP sequences, which was driven by a short CAG promoter (Figure 1A and B, see Materials and ethods).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Then, five rAAV2-retro viruses with different barcodes were individually injected into five brain regions of the same mouse, including AI, DMS, BLA, LH, and MD.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	These brain areas are the known downstream brain regions of vmPFC (Hunnicutt et al., 2016; Hurley et al., 1991; Reppucci and Petrovich, 2016; Vertes, 2004; Zhu et al., 2020).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	A period of six weeks was set to allow efficient retrograde labeling of vmPFC neurons by these barcoded viruses.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	These mice were then sacrificed and the vmPFC (specifically the prelimbic area [PrL] and the infralimbic area [IL]) was carefully dissected for scRNA-seq analysis (Figure 1A).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Single-cell transcriptional libraries were obtained using 10x Genomics library preparation protocols, and virus barcode expression libraries were obtained using user-defined primers, which could enrich cDNA fragments composed of barcode index, unique molecular identifiers (UMIs), and the cell barcode (Figure 1B).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	We detected 24,788 cells in the raw data matrix.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Following initial quality control, which ensured the number of detected RNA in each cell ranged between 500 and 8000, RNA UMI counts in each cell were within 1000–60,000, and the percentage of mitochondrial genes remained below 20%, we recovered 1791 cells undergoing fluorescence-activated cell sorting (FACS) from three mice and 19,470 single cells without sorting from the other three mice, a total of 21,261 cells.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Transcriptional profiling of all cells revealed major cell types including excitatory neurons (Slc17a7), microglia (C1qa), endothelial cells (Itm2a, Endo), oligodendrocyte progenitor cells (Olig2Mog, OPCs), oligodendrocyte (Olig2Mog, Oligo), inhibitory neurons (Gad1), astrocyte (Aldh1l1, Astro), and activated microglia (C1qaPf4, Act.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Microglia) as previously reported (Bhattacherjee et al., 2019; Figure 1C–E).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Barcoded cells below refer to a collection of barcoded cells from unsorted group and FAC-sorted group. (
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	A) Schematic diagram of the experimental workflow. (
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	B) rAAV2 plasmid vector design, and schematic of designed primers to recover cell barcode and UMI in read 1, and 3’ tail of EGFP and virus barcode in read 2.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	According to the recommendation of 10x Genomics, a faithful mapping should cover 28 bp for read 1 and 91 bp for read 2.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	In our design, 150 bp pair-end sequencing can sufficiently meet the need to recover cell barcode, UMI and virus barcode. (
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	C) Umap embedding of transcriptional clustering results for all vmPFC cells. (
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	D) Stacked violin plots showing the expression of markers for each cluster. (
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	E) Heatmap showing the gene-expression correlation between major cell types defined by scRNA-seq of this study and Bhattacherjee et al., 2019. (
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	F) Umap embedding of all determined barcoded cells labeled in blue. (
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	G) Bar plot showing frequency of barcoded (blue) and non-barcoded (grey) cells in all recovered cell types.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	In (C–E), 21,261cells were represented.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	In (F, G), 20,047cells were represented.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	1214 cells with exceptionally high nUMIs were removed.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Figure 1—figure supplement 1.Validation of rAAV2-retro injection sites and determination of valid barcoded cells.(A) The position of injection sites (AI-1, AI-2, DMS, LH, MD, BLA) to deliver rAAV2-retro-EGFP plotted on coronal section diagrams (top).
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	The left corner values indicate the anteroposterior distance of the section from bregma.
PMC10914349	High-throughput mapping of single-neuron projection and molecular features by retrograde barcoded labeling.	Representative immunohistochemistry images showing rAAV2-retro-EGFP injection sites in coronal sections (bottom).

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

For projecting to HNOCA, the query model was based on the scPoli model pretrained with the HNOCA data, and fineturned with a batch size of 16,384 for a maximum of 30 epochs with 20 pretraining epochs.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

A nearest neighbour graph was created for the disease-modelling atlas on the basis of the projected latent representation to HNOCA with scanpy (default parameters), with which a UMAP embedding was created with scanpy (default parameters).

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Next, for both HNOCA and the disease-modelling atlas, cells were represented by the concatenated representation of HNOCA-scPoli and primary-scANVI models.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

A bipartite wkNN graph was then reconstructed as mentioned above, by identifying 50 nearest neighbours in HNOCA for each disease-modelling atlas cell.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

On the basis of the bipartite wkNN, the majority voting-based label transfer was applied to transfer the four levels of hierarchical cell type annotation and regional identity to the disease-modelling atlas.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

For each cell in the disease-modelling atlas, a matched HNOCA metacell was reconstructed on the basis of the above mentioned bipartite wkNN.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

In brief, for a query cell i and a gene j measured in HNOCA, its matched metacell expression of j, denoted as , is calculated as:[12pt] $$_}^ }=__}_}_}}__}_}}$$′=∑k⊆Niwikekj∑k⊆NiwikHere, Ni represents all HNOCA nearest neighbours of the query cell ci, wik represents the edge weight between query cell i and reference cell k, and ekj represents expression level of gene j in reference cell k. Given the matched HNOCA metacell transcriptomic profile, the similarity between a query cell and its matched cell state in HNOCA is then calculated as the Spearman correlation between the query cell transcriptomic profile and its matched HNOCA metacell transcriptomic profile.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

To analyse the glioblastoma organoid dataset (GBM-2019), cells from the publication were subset from the integrated disease-modelling atlas.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Using scanpy, highly variable genes were identified with default parameters.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

The log-normalized expression values of the highly variable genes were then scaled across cells, the truncated PCA was performed with the top 20 principal components used for the following analysis.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Next, harmonypy, the Python implementation of harmony, was applied to integrate cells from different samples.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

On the basis of the harmony-integrated embeddings, the neighbour graph was reconstructed.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

UMAP embeddings and Louvain clusters (resolution, 0.5) were created on the basis of the nearest neighbour graph.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Among the 12 identified clusters, cluster-7 and cluster-0, the two clusters with the highest AQP4 expression, were selected for the following DE analysis.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

To analyse the FXS dataset (FXS-2021), cells from the publication were subset from the integrated disease-modelling atlas.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

The same procedure of highly variable gene identification, data scaling and PCA as the GBM-2019 dataset was applied.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Next, the nearest neighbour graph was created directly on the basis of the top 20 principal components.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

UMAP embeddings and Louvain clusters (resolution, 1) were then created on the basis of the reconstructed nearest neighbour graph.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Among the 30 clusters, cluster-17 and cluster-23, which express EMX1 and FOXG1 and were largely predicted to be dorsal telencephalic NPCs and neurons according to the transferred labels from HNOCA, were selected for the following DE analysis.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

To compare expression levels of two groups of paired cells, the expression difference per gene of each cell pair is first calculated on the basis of the log-normalized expression values.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Next, for each gene to test for DE, its variance over the calculated expression difference per cell pair (σ) is compared with the sum of squared of expression differences (di for gene i) normalized by the number of cell pairs:[12pt] $$_^=_^_}.$$02=∑i=1ndin.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Here, an F-test is applied for the comparison, with f = σ/s0, d.f.1 = n − 1 and d.f.2 = n. To construct the HNOCA-CE, we first collected raw count matrices and associated metadata of five more neural organoid studies.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

For two publications, we obtained them from the sources listed in the ‘Data availability’ section of the paper.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

For the remaining three publications, count matrices and associated metadata were provided directly by the authors.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

We subset each dataset to the healthy control cells and removed any cells with fewer than 200 genes expressed.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

We subset the gene space of every dataset to the 3,000 HVGs of HNOCA while filling the expression of missing genes in the community datasets with zeros.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

On average, 23% of genes with zero expression were added per dataset.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

We instantiated a mapping object from the HNOCA-tools package (at commit fe38c52) using the saved scPoli model weights from the HNOCA integration.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Using the map_query method of the mapper instance, we projected the community datasets to HNOCA.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

We used the following training hyperparameters: retrain = ‘partial’, batch_size = 256, unlabeled_prototype_training = False, n_epochs = 10, pretraining_epochs = 9, early_stopping_kwargs = early_stopping_kwargs, eta = 10, alpha_epoch_anneal = 10.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

We computed the wkNN graph using the compute_wknn method of the mapper instance with k = 100.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

We transferred the final level_2 cell type labels from HNOCA to the community datasets using this neighbour graph.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

To obtain the combined representation of HNOCA-CE, we projected HNOCA together with the added community datasets through the trained model and computed a neighbour graph and UMAP from the resulting latent representation.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

PMC11578878

An integrated transcriptomic cell atlas of human neural organoids.

Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at 10.1038/s41586-024-08172-8.

PMC10914349