"import_samples_monocle" -

[plhm]

Hi, J0bbie.

A couple of things going on with 'import_samples_monocle."
The function crashed on me when I ran the following command on R version 4.3.1.
Here is the command:

import_samples_monocle(folder = "/path-to/alignment", samples ="rerio", gtf = /path/to/Danio.gtf")

And got the following error:

INFO [2024-01-02 15:24:23] import_samples_monocle: Retrieving files.
INFO [2024-01-02 15:24:23] import_samples_monocle: Retrieving and processing GTF file.
Error in `dplyr::filter()`:
ℹ In argument: `base::grepl("protein_coding|lncRNA|IG_.*_gene", gene_type)`.
Caused by error:
! object 'gene_type' not found
Run `rlang::last_trace()` to see where the error occurred.

When I dug into the code, I noted the following problems this far:

The user is not toldthey need to be in the folder containing "Solo.out."
Function expects foder name within Solo.out to be named "GeneFull/filtered," which breaks using the latest version of sci-seq the pipeline, which produces 'GeneFull_Ex50pAS' folders.
Rather than 'gene_type' NCBI uses the column name 'gene_biotype.'
Rather than 'gene_name' NCBI uses the column name 'gene_synonym.'

[plhm]

I tweaked the function a bit, and this seems to be working for me, with the current version of sci-rocket (as of 2024-01-02):

function (folder, samples, gtf, metadata = NULL)
{
    checkmate::assertAccess(folder, access = "r")
    checkmate::assertCharacter(samples)
    checkmate::assertFileExists(gtf, access = "r")
    checkmate::assertTibble(metadata, null.ok = TRUE)
    if (!is.null(metadata) & is.null(metadata$sample_name))
        stop("metadata should contain a column: sample_name")
    futile.logger::flog.info("import_samples_monocle: Retrieving files.")
    files <- tibble::tibble(path = list.files(folder, pattern = "barcodes_converted.tsv$|features.tsv$|matrix.mtx$",
        recursive = TRUE, full.names = TRUE))
    files <- files %>% dplyr::filter(base::grepl("GeneFull_Ex50pAS/filtered/",
        path)) %>% dplyr::mutate(sample = base::sub("_Solo.out.*",
        "", base::sub("/$|^/", "", base::sub(folder, "", path))),
        sample = base::sub("_[^_]*$", "", sample), file = base::basename(path)) %>%
        dplyr::filter(sample %in% samples)
    futile.logger::flog.info("import_samples_monocle: Retrieving and processing GTF file.")
    data_gtf <- rtracklayer::import(gtf, format = "gff") %>%
        tibble::as_tibble() %>% dplyr::filter(type == "gene",
        !base::is.na(gene_id)) %>% dplyr::filter(base::grepl("protein_coding|lncRNA|IG_.*_gene",
        gene_biotype)) %>% dplyr::filter(!base::grepl("^Gm[0-9]",
        gene_synonym) | gene_synonym == "Gm2a") %>% dplyr::distinct(seqnames,
        start, end, strand, gene_id, gene_biotype, gene_synonym) %>%
        dplyr::select(gene_chr = seqnames, gene_start = start,
            gene_end = end, gene_strand = strand, gene_id, gene_biotype,
            gene_synonym)
    futile.logger::flog.info("import_samples_monocle: Generating cell_data_set object for %s samples",
        dplyr::n_distinct(files$sample))
    cds_samples <- future.apply::future_lapply(unique(files$sample),
        function(current_sample) {
            cds <- monocle3::load_mm_data(mat_path = files %>%
                dplyr::filter(sample == current_sample, file ==
                  "matrix.mtx") %>% dplyr::pull(path), feature_anno_path = files %>%
                dplyr::filter(sample == current_sample, file ==
                  "features.tsv") %>% dplyr::pull(path), cell_anno_path = files %>%
                dplyr::filter(sample == current_sample, file ==
                  "barcodes_converted.tsv") %>% dplyr::pull(path),
                feature_metadata_column_names = c("gene_short_name",
                  "type"), umi_cutoff = 0)
            cds <- cds[SummarizedExperiment::rowData(cds)$gene_short_name %in%
                data_gtf$gene_id, ]
            monocle3::fData(cds) <- monocle3::fData(cds) %>%
                tibble::as_tibble(rownames = "gene_id") %>% dplyr::left_join(data_gtf,
                by = "gene_id") %>% tibble::column_to_rownames("gene_id") %>%
                S4Vectors::DataFrame()
            SummarizedExperiment::colData(cds)$sample_name <- base::factor(current_sample)
            return(cds)
        }, future.seed = TRUE)
    base::names(cds_samples) <- base::vapply(cds_samples, FUN = function(x) {
        base::levels(monocle3::pData(x)$sample_name)
    }, "")
    futile.logger::flog.info("import_samples_monocle: Combining into a single cell_data_set.")
    cds_combined <- monocle3::combine_cds(cds_samples, keep_all_genes = FALSE)
    SummarizedExperiment::colData(cds_combined)$sample_name <- NULL
    if (!is.null(metadata)) {
        futile.logger::flog.info("import_samples_monocle: Adding metadata.")
        monocle3::pData(cds_combined) <- monocle3::pData(cds_combined) %>%
            tibble::as_tibble(rownames = "cell_id") %>% dplyr::left_join(metadata,
            by = c(sample = "sample_name")) %>% tibble::column_to_rownames("cell_id") %>%
            S4Vectors::DataFrame()
    }
    monocle3::fData(cds_combined) <- S4Vectors::cbind(data.frame(id = base::gsub("\\..*",
        "", S4Vectors::rownames(cds_combined))), monocle3::fData(cds_combined))
    cds_combined$log.n.umi <- log10(cds_combined$n.umi)
    return(cds_combined)
}