Dataset Information

Multimodal hierarchical classification allows for efficient annotation of CITE-seq data

ABSTRACT: Single-cell RNA sequencing (scRNA-seq) is an invaluable tool for profiling cells in complex tissues and dissecting activation states that lack well-defined surface protein expression. For immune cells, the transcriptomic profile captured by scRNA- seq cannot always identify cell states and subsets defined by conventional flow cytometry. Emerging technologies have enabled multimodal sequencing of single cells, such as paired sequencing of the transcriptome and surface proteome by CITE-seq, but integrating these high dimensional modalities for accurate cell type annotation remains a challenge in the field. Here, we describe a machine learning tool called MultiModal Classifier Hierarchy (MMoCHi) for the cell-type annotation of CITE-seq data. Our classifier involves several steps: 1) we use landmark registration to remove batch-related staining artifacts in CITE-Seq protein expression, 2) the user defines a hierarchy of classifications based on cell type similarity and ontology and provides markers (protein or gene expression) for the identification of ground truth populations within the dataset by threshold gating, 3) progressing through this user-defined hierarchy, we train a random forest classifier using all available modalities (surface proteome and transcriptome data), and 4) we use these forests to predict cell types across the entire dataset. Applying MMoCHi to CITE-seq data of immune cells isolated from eight distinct tissue sites of two human organ donors yields high-purity cell type annotations encompassing the broad array of immune cell states in the dataset. This includes T and B cell memory subsets, macrophages and monocytes, and natural killer cells, as well as rare populations of plasmacytoid dendritic cells, innate T cells, and innate lymphoid cell subsets. We validate the use of feature importances extracted from the classifier hierarchy to select robust genes for improved identification of T cell memory subsets by scRNA-seq. Together, MMoCHi provides a comprehensive system of tools for the batch-correction and cell- type annotation of CITE-seq data. Moreover, this tool provides flexibility in classification hierarchy design allowing for cell type annotations to reflect a researcher’s specific experimental design. This flexibility also renders MMoCHi readily extendable beyond immune cell annotation, and potentially adaptable to other sequencing modalities.

ORGANISM(S): Homo sapiens

PROVIDER: GSE229791 | GEO | 2023/08/08

REPOSITORIES: GEO

ACCESS DATA

Similar Datasets

Project description:This repository contains all the FASTQ files for the five data modalities (scRNA-seq, scATAC-seq, Multiome, CITE-seq+scVDJ-seq, and spatial transcriptomics) used in the article \\"An Atlas of Cells in The Human Tonsil,\\" published in Immunity in 2024. Inspired by the TCGA barcodes, we have named each fastq file with the following convention: [TECHNOLOGY].[DONOR_ID].[SUBPROJECT].[GEM_ID].[LIBRARY_ID].[LIBRARY_TYPE].[LANE].[READ].fastq.gz which allows to retrieve all metadata from the name itself. Here is a full description of each field: - TECHNOLOGY: scRNA-seq, scATAC-seq, Multiome, CITE-seq+scVDJ-seq, and spatial transcriptomics (Visium). We also include the fastq files associated with the multiome experiments performed on two mantle cell lymphoma patients (MCL). - DONOR_ID: identifier for each of the 17 patients included in the cohort. We provide the donor-level metadata in the file \\"tonsil_atlas_donor_metadata.csv\\", including the hospital, sex, age, age group, cause for tonsillectomy and cohort type for every donor. - SUBPROJECT: each subproject corresponds to one run of the 10x Genomics Chromium™ Chip. - GEM_ID: each run of the 10x Genomics Chromium™ Chip consists of up to 8 \\"GEM wells\\" (see https://www.10xgenomics.com/support/software/cell-ranger/getting-started/cr-glossary): a set of partitioned cells (Gel Beads-in-emulsion) from a single 10x Genomics Chromium™ Chip channel. We give a unique identifier to each of these channels. - LIBRARY_ID: one or more sequencing libraries can be derived from a GEM well. For instance, multiome yields two libraries (ATAC and RNA) and CITE-seq+scVDJ yields 4 libraries (RNA, ADT, BCR, TCR). - LIBRARY_TYPE: the type of library for each library_id. Note that we used cell hashing () for a subset of the scRNA-seq libraries, and thus the library_type can be \\"not_hashed\\", \\"hashed_cdna\\" (RNA expression) or \\"hashed_hto\\" (the hashtag oligonucleotides). - LANE: to increase sequencing depth, each library was sequenced in more than one lane. Important: all lanes corresponding to the same sequencing library need to be inputed together to cellranger, because they come from the same set of cells. - READ: for scATAC-seq we have three reads (R1, R2 or R3), see cellranger-atac's documentation. While we find these names to be the most useful, they need to be changed to follow cellranger's conventions. We provide a code snippet in the README file of the GitHub repository associated with the tonsil atlas to convert between both formats (https://github.com/Single-Cell-Genomics-Group-CNAG-CRG/TonsilAtlas/). Besides the fastq files, cellranger (and other mappers) require additional files, which we also provide in this repository: - cell_hashing_metadata.csv: as mentioned above, we ran cell hashing (10.1186/s13059-018-1603-1) to detect doublets and reduce cost per cell. This file provides the sequence of the hashtag oligonucleotides in cellranger convention to allow demultiplexing. - cite_seq_feature_reference.csv: similar to the previous file, this one links each protein surface marker to the hashtag oligonucleotide that identified it in the CITE-seq experiment. - V10M16-059.gpr and V19S23-039.gpr: these correspond to the two slides of the two Visium experiments performed in the tonsil atlas. They are needed to run spaceranger. - [GEM_ID]_[SLIDE]_[CAPTURE_AREA].jpg: 8 images associated with the Visium experiments. Here, GEM_ID refers to each of the 4 capture areas in each slide. - [TECHNOLOGY]_sequencing_metadata.csv: the GEM-level metadata for each technology. It includes the relationship between subproject, gem_id, library_id, library_type and donor_id. These are the other repositories associated with the tonsil atlas: - Expression and accessibility matrices: https://zenodo.org/records/10373041 - Seurat objects: https://zenodo.org/records/8373756 - HCATonsilData package: https://bioconductor.org/packages/release/data/experiment/html/HCATonsilData.html - Azimuth: https://azimuth.hubmapconsortium.org/ - Github: https://github.com/Single-Cell-Genomics-Group-CNAG-CRG/TonsilAtlas

Dataset Information

Multimodal hierarchical classification allows for efficient annotation of CITE-seq data

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure