R/seurat_annotation.R

Defines functions annotate_excluded add_percent_mito add_read_count_col transcripts_to_genes genes_to_transcripts annotate_cell_cycle

Documented in add_percent_mito add_read_count_col annotate_cell_cycle annotate_excluded genes_to_transcripts transcripts_to_genes

#' Annotate Cell Cycle
#'
#' Annotate Cell Cycle for Gene and Transcript Seurat Objects
#'
#' @param seu A seurat object
#' @param organism organism. Default "human"
#'
#' @return
#' @export
#'
#' @examples
#' annotate_cell_cycle(panc8, organism = "human")
#' annotate_cell_cycle(baron2016singlecell, organism = "mouse")
annotate_cell_cycle <- function(seu, organism = "human", ...) {
    # setdefaultassay to "gene"
    # Seurat::DefaultAssay(seu) <- "gene"

    s_genes <- cc.genes$s.genes
    g2m_genes <- cc.genes$g2m.genes

    if (organism == "mouse") {
        s_genes <-
            dplyr::filter(human_to_mouse_homologs, HGNC.symbol %in% s_genes) %>%
            dplyr::pull(MGI.symbol)

        g2m_genes <-
            dplyr::filter(human_to_mouse_homologs, HGNC.symbol %in% g2m_genes) %>%
            dplyr::pull(MGI.symbol)
    }

    seu <- CellCycleScoring(seu, s.features = s_genes, g2m.features = g2m_genes, set.ident = FALSE)
}

#' Gene Symbols to Ensembl Transcript Ids
#'
#' convert hgnc gene symbols to ensembl transcript ids
#'
#' @param genes
#'
#' @return
#' @export
#'
#' @examples
#'
#' genes_to_transcripts("RXRG")
#'
genes_to_transcripts <- function(genes, organism = "human") {
    if (organism == "human") {
        feature_table <- ensembldb::transcripts(EnsDb.Hsapiens.v86::EnsDb.Hsapiens.v86,
            columns = c("gene_name", "gene_biotype", "gene_id"),
            return.type = "DataFrame"
        )
    } else if (organism == "mouse") {
        feature_table <- ensembldb::transcripts(EnsDb.Mmusculus.v79::EnsDb.Mmusculus.v79,
            columns = c("gene_name", "gene_biotype", "gene_id"),
            return.type = "DataFrame"
        )
    }

    feature_table %>%
        as_tibble() %>%
        dplyr::filter(gene_name %in% genes) %>%
        dplyr::pull(tx_id)
}

#' Ensembl Transcript Ids to Gene Symbols
#'
#' convert ensembl transcript ids to hgnc gene symbols
#'
#' @param transcripts
#'
#' @return
#' @export
#'
#' @examples
#'
#' RXRG_transcripts_hs <- c("ENST00000359842", "ENST00000470566", "ENST00000465764", "ENST00000619224")
#'
#' transcripts_to_genes(transcripts = RXRG_transcripts_hs)
#'
transcripts_to_genes <- function(transcripts, organism = "human") {
    if (organism == "human") {
        gene_table <- annotables::grch38
        transcript_table <- annotables::grch38_tx2gene
    } else if (organism == "mouse") {
        gene_table <- annotables::grcm38
        transcript_table <- annotables::grcm38_tx2gene
    }

    tibble::tibble(enstxp = transcripts) %>%
        dplyr::left_join(transcript_table, by = "enstxp") %>%
        dplyr::left_join(gene_table, by = "ensgene") %>%
        dplyr::pull("symbol") %>%
        identity()
}

#' Annotate Low Read Count Category
#'
#' Add a Read Count Categorical Variable to Seurat Object (based on nCount_RNA)
#'
#' @param seu A seurat object
#' @param thresh Set a threshold for low read count
#'
#' @return
#' @export
#'
#' @examples
add_read_count_col <- function(seu, thresh = 1e5) {
    rc <- seu[["nCount_gene"]] < thresh

    seu <- Seurat::AddMetaData(
        object = seu,
        metadata = rc,
        col.name = "low_read_count"
    )
}

#' Annotate percent mitochondrial reads per cell
#'
#'  Add a Read Count Categorical Variable to Seurat Object (based on nCount_RNA)
#'
#' @param seu A seurat object
#'
#' @param seurat_assay gene
#'
#' @return
#' @export
#'
#' @examples
#' add_percent_mito(panc8)
#' add_percent_mito(baron2016singlecell, organism = "mouse")
add_percent_mito <- function(seu, organism = organism, seurat_assay = "gene") {

    seu[["percent.mt"]] <- PercentageFeatureSet(seu, pattern = "^MT-", assay = seurat_assay)

    return(seu)
}


#' Annotate Exclusion Criteria
#'
#' @param seu A seurat object
#' @param excluded_cells a named list of cells to be excluded of the form list(set1 = c(cell1, celll2), set2 = c(cell3, cell4))
#' all other cells will be marked NA in a column titled "excluded_because"
#'
#' @return
#' @export
#'
#' @examples
annotate_excluded <- function(seu, ...) {
    # consider replacing
    # mutate(coalesce_var = coalesce(!!! syms(vars_select(names(.), dplyr::starts_with("my")))))

    excluded_cells <- list(...)

    excluded_cells <- purrr::map2(excluded_cells, names(excluded_cells), ~ rep(.y, length(.x))) %>%
        unlist() %>%
        purrr::set_names(unlist(excluded_cells)) %>%
        tibble::enframe("sample_id", "excluded_because")

    excluded_because <- tibble::as_tibble(seu[["nCount_RNA"]], rownames = "sample_id") %>%
        dplyr::full_join(excluded_cells, by = "sample_id")

    if ("excluded_because.x" %in% colnames(excluded_because)) {
        excluded_because <- dplyr::coalesce(excluded_because, excluded_because.x, excluded_because.y) %>%
            dplyr::select(-nCount_RNA) %>%
            tibble::deframe() %>%
            identity()
    } else {
        excluded_because <- select(excluded_because, -nCount_RNA) %>%
            tibble::deframe() %>%
            identity()
    }

    seu <- Seurat::AddMetaData(
        object = seu,
        metadata = excluded_because,
        col.name = "excluded_because"
    )

    return(seu)
}
whtns/seuratTools documentation built on Oct. 28, 2024, 7:18 a.m.