biopipen 0.17.7__py3-none-any.whl → 0.18.0__py3-none-any.whl

biopipen/scripts/scrna/SeuratClusterStats.R

@@ -1,535 +1,21 @@
 source("{{biopipen_dir}}/utils/misc.R")
 source("{{biopipen_dir}}/utils/plot.R")
+library(jsonlite)
+library(slugify)
 library(Seurat)
 library(rlang)
 library(dplyr)
 library(tibble)
 library(ggprism)
+library(ggsci)
+library(ggrepel)
 library(tidyseurat)
 
 srtfile = {{in.srtobj | r}}
 outdir = {{out.outdir | r}}
-envs = {{envs | r: todot="-"}}
 
 srtobj = readRDS(srtfile)
 
-do_stats_cells = function(casename, devpars, odir, by = NULL, frac = FALSE, filtering = NULL) {
-    plotfile = file.path(odir, paste0(casename, ".png"))
-    txtfile = paste0(plotfile, ".txt")
-    if (is.null(devpars)) {
-        devpars = list(res = 100, height = 1000)
-    } else {
-        devpars$res = if (is.null(devpars$res)) default_devpars$res else devpars$res
-        devpars$height = if (is.null(devpars$height)) default_devpars$height else devpars$height
-    }
-    if (frac) {
-        ylab = paste("Fraction of cells")
-        mapping_y = "cellFraction"
-    } else {
-        ylab = paste("Number of cells")
-        mapping_y = "nCells"
-    }
-    df_cells = srtobj@meta.data
-    if (!is.null(filtering)) {
-        df_cells = df_cells %>% filter(!!rlang::parse_expr(filtering))
-    }
-    if (is.null(by)) {
-        df_cells = df_cells %>%
-            group_by(seurat_clusters) %>%
-            summarize(nCells = n(), .groups = "keep") %>%
-            mutate(cellFraction = nCells / sum(nCells))
-
-        if (is.null(devpars$width)) { devpars$width = 1000 }
-        plotGG(
-            df_cells,
-            geom = "col",
-            args = list(
-                mapping = aes_string(x="seurat_clusters", y=mapping_y)
-            ),
-            ggs = c(
-                paste0('ggtitle("', ylab, ' for each cluster")'),
-                'theme_prism()',
-                'theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))',
-                paste0('labs(x="Seurat Cluster", y="', ylab, '")')
-            ),
-            devpars = devpars,
-            outfile = plotfile
-        )
-    } else {
-        df_cells = df_cells %>%
-            group_by(!!sym(by), seurat_clusters) %>%
-            summarize(nCells = n()) %>%
-            group_by(seurat_clusters) %>%
-            mutate(cellFraction = nCells / sum(nCells))
-
-        if (is.null(devpars$width)) {
-            devpars$width = 800 + 200 * ceiling(length(unique(df_cells[[by]])) / 20)
-        }
-        plotGG(
-            df_cells,
-            geom = "col",
-            args = list(
-                mapping = aes_string(x="seurat_clusters", y=mapping_y, fill=by),
-                position = "stack"
-            ),
-            ggs = c(
-                paste0('ggtitle("', ylab, ' for each cluster by ', by, '")'),
-                'theme_prism()',
-                'theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))',
-                paste0('labs(x="Seurat Cluster", y="', ylab, '")')
-            ),
-            devpars = devpars,
-            outfile = plotfile
-        )
-    }
-
-    write.table(
-        df_cells,
-        file = txtfile,
-        quote = FALSE,
-        sep = "\t",
-        row.names = FALSE,
-        col.names = TRUE
-    )
-}
-
-do_stats = function() {
-    odir = file.path(outdir, "stats")
-    dir.create(odir, showWarnings = FALSE)
-    for (name in names(envs$stats)) {
-        stat_pars = envs$stats[[name]]
-        args = list(
-            casename = name,
-            devpars = stat_pars$devpars,
-            odir = odir,
-            by = stat_pars$by,
-            frac = FALSE,
-            filtering = stat_pars$filter
-        )
-        if (startsWith(name, "fracCells")) {
-            args$frac = TRUE
-        } else if (!startsWith(name, "nCells")) {
-            warning(paste("Unknown stat:", name, ", skipping"))
-            next
-        }
-
-        do_call(do_stats_cells, args)
-    }
-}
-
-.get_outfile = function(odir, prefix, ext = "png") {
-    i = 1
-    while (TRUE) {
-        outfile = file.path(odir, paste0(prefix, "-", i, ".", ext))
-        if (!file.exists(outfile)) {
-            return(outfile)
-        }
-        i = i + 1
-    }
-    return(outfile)
-}
-
-.get_features = function(features, default_features, default = NULL) {
-    if (is.null(default)) {
-        default = VariableFeatures(srtobj)
-    }
-    # When nothing passed, use the genes
-    if (is.null(features)) {
-        if (is.null(default_features)) {
-            return (default[1:20])
-        } else {
-            return (default_features)
-        }
-    }
-    # When multiple items passed, use them as features
-    if (length(features) > 1) {
-        return (features)
-    }
-    # See if it is "default"
-    if (is.integer(features)) {
-        return (default[1:features])
-    }
-    # See if it is a file
-    if (!file.exists(features)) {
-        return (features)
-    }
-    # length(features) == 1 && file.exists(features[1])
-    feats = read.table(features, header = FALSE, row.names = NULL, check.names = FALSE)
-    feats$V1
-}
-
-do_feats_ridgeplots = function(odir, pms, default_features) {
-    outfile = .get_outfile(odir, "ridgeplots")
-
-    devpars = pms$devpars
-    pms$devpars = NULL
-    plus = pms$plus
-    pms$plus = NULL
-    title = pms$title
-    pms$title = NULL
-    if (is.null(title)) {
-        title = tools::file_path_sans_ext(basename(outfile))
-    }
-    cat(title, file = paste0(outfile, ".title"))
-    subsetpms = pms$subset
-    pms$subset = NULL
-    if (is.null(plus)) {
-        plus = c()
-    }
-    pms$features = .get_features(pms$features, default_features)
-    if (is.null(pms$ncol)) {
-        pms$ncol = min(2, length(pms$features))
-    }
-    if (is.null(devpars)) {
-        devpars = list(
-            width = 400 * pms$ncol,
-            height = ceiling(length(pms$features) / pms$ncol) * 400,
-            res = 100
-        )
-    }
-    if (is.null(subsetpms)) {
-        pms$object = srtobj
-    } else {
-        pms$object = srtobj %>% filter(eval(parse(text=subsetpms)))
-    }
-    p = do_call(RidgePlot, pms)
-    for (pls in plus) {
-        p = p + eval(parse(text = pls))
-    }
-    devpars$filename = outfile
-    do_call(png, devpars)
-    print(p)
-    dev.off()
-}
-
-
-do_feats_vlnplots = function(odir, pms, default_features) {
-    outfile = .get_outfile(odir, "vlnplots")
-
-    devpars = pms$devpars
-    pms$devpars = NULL
-    boxplot = pms$boxplot
-    pms$boxplot = NULL
-    plus = pms$plus
-    pms$plus = NULL
-    subsetpms = pms$subset
-    pms$subset = NULL
-    title = pms$title
-    pms$title = NULL
-    if (is.null(title)) {
-        title = tools::file_path_sans_ext(basename(outfile))
-    }
-    cat(title, file = paste0(outfile, ".title"))
-    if (is.null(plus)) {
-        plus = c()
-    }
-    if (!is.null(boxplot) && length(boxplot) == 0) {
-        boxplot = list(width = .1, fill = "white")
-    }
-    pms$features = .get_features(pms$features, default_features)
-    if (is.null(pms$ncol)) {
-        pms$ncol = min(2, length(pms$features))
-    }
-    if (is.null(devpars)) {
-        devpars = list(
-            width = 400 * pms$ncol,
-            height = ceiling(length(pms$features) / pms$ncol) * 400,
-            res = 100
-        )
-    }
-    if (is.null(subsetpms)) {
-        pms$object = srtobj
-    } else {
-        pms$object = srtobj %>% tidyseurat::filter(eval(parse(text=subsetpms)))
-    }
-    p = do_call(VlnPlot, pms)
-    if (!is.null(boxplot)) {
-        p = p + do_call(geom_boxplot, boxplot)
-    }
-    for (pls in plus) {
-        p = p + eval(parse(text = pls))
-    }
-    devpars$filename = outfile
-    do_call(png, devpars)
-    print(p)
-    dev.off()
-}
-
-
-do_feats_featureplots = function(odir, pms, default_features) {
-    outfile = .get_outfile(odir, "featureplots")
-
-    devpars = pms$devpars
-    pms$devpars = NULL
-    title = pms$title
-    pms$title = NULL
-    if (is.null(title)) {
-        title = tools::file_path_sans_ext(basename(outfile))
-    }
-    cat(title, file = paste0(outfile, ".title"))
-    pms$features = .get_features(pms$features, default_features)
-    if (is.null(pms$ncol)) {
-        pms$ncol = min(2, length(pms$features))
-    }
-    if (is.null(devpars)) {
-        devpars = list(
-            width = 400 * pms$ncol,
-            height = ceiling(length(pms$features) / pms$ncol) * 300,
-            res = 100
-        )
-    }
-    subsetpms = pms$subset
-    pms$subset = NULL
-    if (is.null(subsetpms)) {
-        pms$object = srtobj
-    } else {
-        pms$object = srtobj %>% filter(eval(parse(text=subsetpms)))
-    }
-    p = do_call(FeaturePlot, pms)
-    devpars$filename = outfile
-
-    tryCatch({
-        do_call(png, devpars)
-        print(p)
-        dev.off()
-    }, error = function(e) {
-        stop(
-            paste(
-                paste(names(devpars), collapse=" "),
-                paste(devpars, collapse=" "),
-                e,
-                sep = "\n"
-            )
-        )
-    })
-}
-
-do_feats_dotplot = function(odir, pms, default_features) {
-    outfile = .get_outfile(odir, "dotplot")
-
-    devpars = pms$devpars
-    pms$devpars = NULL
-    plus = pms$plus
-    pms$plus = NULL
-    subsetpms = pms$subset
-    pms$subset = NULL
-    title = pms$title
-    pms$title = NULL
-    if (is.null(title)) {
-        title = tools::file_path_sans_ext(basename(outfile))
-    }
-    cat(title, file = paste0(outfile, ".title"))
-    if (is.null(plus)) {
-        plus = c()
-    }
-    pms$features = .get_features(pms$features, default_features)
-    if (is.null(devpars)) {
-        devpars = list(
-            height = length(unique(srtobj@meta.data$seurat_clusters)) * 80 + 150,
-            width = length(pms$features) * 50 + 150,
-            res = 100
-        )
-    }
-    if (is.null(subsetpms)) {
-        pms$object = srtobj
-    } else {
-        pms$object = srtobj %>% filter(eval(parse(text=subsetpms)))
-    }
-    p = do_call(DotPlot, pms)
-    for (pls in plus) {
-        p = p + eval(parse(text = pls))
-    }
-    devpars$filename = outfile
-    tryCatch({
-        do_call(png, devpars)
-        print(p)
-        dev.off()
-    }, error = function(e) {
-        stop(
-            paste(
-                paste(names(devpars), collapse=" "),
-                paste(devpars, collapse=" "),
-                e,
-                sep = "\n"
-            )
-        )
-    })
-}
-
-
-do_feats_heatmap = function(odir, pms, default_features) {
-    outfile = .get_outfile(odir, "heatmap")
-
-    devpars = pms$devpars
-    pms$devpars = NULL
-    plus = pms$plus
-    pms$plus = NULL
-    subsetpms = pms$subset
-    pms$subset = NULL
-    title = pms$title
-    pms$title = NULL
-    if (is.null(title)) {
-        title = tools::file_path_sans_ext(basename(outfile))
-    }
-    cat(title, file = paste0(outfile, ".title"))
-    if (is.null(plus)) {
-        plus = c()
-    }
-    pms$features = .get_features(pms$features, default_features)
-    if (is.null(devpars)) {
-        devpars = list(
-            width = length(unique(srtobj@meta.data$seurat_clusters)) * 60 + 150,
-            height = length(pms$features) * 40 + 150,
-            res = 100
-        )
-    }
-    downsample = pms$downsample
-    pms$downsample = NULL
-
-    if (is.null(subsetpms)) {
-        sobj = srtobj
-    } else {
-        sobj = srtobj %>% filter(eval(parse(text=subsetpms)))
-    }
-    if (is.null(downsample)) {
-        pms$object = sobj
-        warn(
-            paste0(
-                "DoHeatmap: `downsample` not specified, using full data. ",
-                "This may cause a blank heatmap. ",
-                "See: https://github.com/satijalab/seurat/issues/2724"
-            ),
-            immediate. = TRUE
-        )
-    } else if (downsample %in% c("average", "mean")) {
-        pms$object = AverageExpression(sobj, return.seurat = TRUE)
-    } else {
-        pms$object = subset(sobj, downsample = downsample)
-    }
-    p = do_call(DoHeatmap, pms)
-    for (pls in plus) {
-        p = p + eval(parse(text = pls))
-    }
-    mapal = colorRampPalette(RColorBrewer::brewer.pal(11,"RdBu"))(256)
-    p = p + scale_fill_gradientn(colours = rev(mapal))
-    devpars$filename = outfile
-    do_call(png, devpars)
-    print(p)
-    dev.off()
-}
-
-
-do_feats_table = function(odir, pms, default_features) {
-    outfile = .get_outfile(odir, "table", "tsv")
-
-    subsetpms = pms$subset
-    log2_scale = pms$log2
-    if (is.null(log2_scale)) { log2_scale = TRUE }
-    features = .get_features(pms$features, default_features)
-    title = pms$title
-    if (is.null(title)) { title = tools::file_path_sans_ext(basename(outfile)) }
-    cat(title, file = paste0(outfile, ".title"))
-
-    if (is.null(subsetpms)) {
-        sobj = srtobj
-    } else {
-        sobj = srtobj %>% filter(eval(parse(text=subsetpms)))
-    }
-    # default slot (data), assay
-    # values are exponentiated prior to averaging so that averaging is done in non-log space.
-    avgdata = AverageExpression(sobj, features = features)
-    edata = avgdata$RNA
-    # replace the missing genes
-    edata[rownames(avgdata$integrated), ] = avgdata$integrated
-    if (log2_scale) { edata = log2(edata) }
-    edata = as.data.frame(edata) %>%
-        rownames_to_column("Gene") %>%
-        select(Gene, everything())
-    write.table(edata, file = outfile, sep = "\t", quote = FALSE, row.names = FALSE, col.names = TRUE)
-}
-
-
-do_feats = function() {
-    if (length(envs$features) == 0) {
-        return (NULL)
-    }
-    odir = file.path(outdir, "features")
-    dir.create(odir, showWarnings = FALSE)
-    default_features = envs$features$features
-    envs$features$features = NULL
-    if (!is.null(default_features) && is.character(default_features) && file.exists(default_features)) {
-        default_features = read.table(default_features, header = FALSE, sep = "\t", row.names = NULL, check.names = FALSE)
-        default_features = default_features[,1,drop=TRUE]
-    } else if (!is.null(default_features) && is.character(default_features)) {
-        default_features = trimws(strsplit(default_features, ",")[[1]])
-    }
-
-    exprplots = names(envs$features)
-    for (name in exprplots) {
-        cat(paste0("Expr plot: ", name, " ...\n"), file = stderr())
-        if (startsWith(name, "ridgeplots")) {
-            do_feats_ridgeplots(odir, envs$features[[name]], default_features)
-        } else if (startsWith(name, "vlnplots")) {
-            do_feats_vlnplots(odir, envs$features[[name]], default_features)
-        } else if (startsWith(name, "featureplots")) {
-            do_feats_featureplots(odir, envs$features[[name]], default_features)
-        } else if (startsWith(name, "dotplot")) {
-            do_feats_dotplot(odir, envs$features[[name]], default_features)
-        } else if (startsWith(name, "heatmap")) {
-            do_feats_heatmap(odir, envs$features[[name]], default_features)
-        } else if (startsWith(name, "table")) {
-            do_feats_table(odir, envs$features[[name]], default_features)
-        } else {
-            print(paste("Unrecognized expression plot type: ", name))
-        }
-    }
-}
-
-do_dimplot = function(odir, dpname, dpenvs) {
-    devpars = dpenvs$devpars
-    dpenvs$devpars = NULL
-    if (is.null(devpars)) {
-        devpars = list(
-            width = 1000,
-            height = 1000,
-            res = 100
-        )
-    }
-    plus = dpenvs$plus
-    dpenvs$plus = NULL
-    if (is.null(plus)) {
-        plus = c()
-    }
-    if (!any(grepl("ggtitle(", plus, fixed=T))) {
-        plus = c(plus, paste0("ggtitle(", dQuote(dpname, q=F), ")"))
-    }
-
-    dpenvs$object = srtobj
-    p = do_call(DimPlot, dpenvs)
-    for (pls in plus) {
-        p = p + eval(parse(text = pls))
-    }
-
-    devpars$filename = file.path(odir, paste0(slugify(dpname), ".png"))
-    do_call(png, devpars)
-    print(p)
-    dev.off()
-}
-
-do_dimplots = function() {
-    if (length(envs$dimplots) == 0) {
-        return (NULL)
-    }
-    odir = file.path(outdir, "dimplots")
-    dir.create(odir, showWarnings = FALSE)
-
-    for (dpname in names(envs$dimplots)) {
-        dpenvs = envs$dimplots[[dpname]]
-        do_dimplot(odir, dpname, dpenvs)
-    }
-}
-
-
-do_stats()
-do_feats()
-do_dimplots()
+{% include biopipen_dir + "/scripts/scrna/SeuratClusterStats-stats.R" %}
+{% include biopipen_dir + "/scripts/scrna/SeuratClusterStats-features.R" %}
+{% include biopipen_dir + "/scripts/scrna/SeuratClusterStats-dimplots.R" %}

biopipen/scripts/scrna/SeuratClustering.R

@@ -114,7 +114,26 @@ id_args = list_setdefault(
     dims = 1:30
 )
 id_args$dims = 1:min(min_dim, max(id_args$dims))
-obj_list = do_call(IntegrateData, id_args)
+tryCatch({
+    obj_list = do_call(IntegrateData, id_args)
+}, error = function(e) {
+    msg = ""
+    if (grepl("number of items to replace is not a multiple of replacement length", e)) {
+        default_kweight = 100
+        if (!is.null(envs$IntegrateData$k.weight)) {
+            default_kweight = envs$IntegrateData$k.weight
+        }
+        msg = paste0(
+            "It's possible that you have too few cells in some samples, ",
+            "causing a small number of anchor cells in the anchorset. \n",
+            "  Try changing `k.weight` for `IntegrateData` by setting ",
+            "`envs.IntegrateData.k-weight` to a smaller number (it's now ",
+            default_kweight, "). \n",
+            "  See also https://github.com/satijalab/seurat/issues/6341"
+        )
+    }
+    stop(paste0(msg, "\n", e))
+})
 
 {%- else -%}
 # ############################

biopipen/scripts/tcr/TESSA.R

@@ -13,6 +13,7 @@ exprfile <- {{in.srtobj | r}}
 outfile <- {{out.outfile | r}}
 python <- {{envs.python | r}}
 within_sample <- {{envs.within_sample | r}}
+assay <- {{envs.assay | r}}
 predefined_b <- {{envs.predefined_b | r}}
 max_iter <- {{envs.max_iter | int}}
 tessa_srcdir <- "{{biopipen_dir}}/scripts/tcr/TESSA_source"
@@ -55,15 +56,18 @@ tcrdata <- do_call(rbind, lapply(seq_len(nrow(immdata$meta)), function(i) {
         mutate(Barcode = glue("{{envs.prefix}}{Barcode}"), sample = Sample)
 }))
 if (has_VJ) {
-    tcrdata <- tcrdata %>% select(
+    tcrdata <- tcrdata %>% dplyr::mutate(
+        v_gene = sub("-\\d+$", "", V.name),
+        j_gene = sub("-\\d+$", "", J.name)
+    ) %>% dplyr::select(
         contig_id = Barcode,
         cdr3 = CDR3.aa,
-        v_gene = V.name,
-        j_gene = J.name,
+        v_gene,
+        j_gene,
         sample
     )
 } else {
-    tcrdata <- tcrdata %>% select(
+    tcrdata <- tcrdata %>% dplyr::select(
         contig_id = Barcode,
         cdr3 = CDR3.aa,
         sample
@@ -77,7 +81,7 @@ is_gz <- endsWith(tolower(exprfile), ".gz")
 
 if (is_seurat) {
     sobj <- readRDS(exprfile)
-    expr <- GetAssayData(sobj)
+    expr <- GetAssayData(sobj, slot = "data", assay = assay)
 } else if (is_gz) {
     expr <- read.table(gzfile(exprfile), sep="\t", header=TRUE, row.names=1)
 } else {

biopipen/scripts/tcr/TESSA_source/BriseisEncoder.py

@@ -59,7 +59,7 @@ def preprocess(filedir):
     dataset = pd.read_csv(filedir, header=0)
     if dataset.isnull().values.any():
         print("Input data contains NAs.")
-        dataset = dataset.dropna()
+        # dataset = dataset.dropna()
     data_new = pd.DataFrame(
         {
             "contig_id": dataset["contig_id"],
@@ -119,7 +119,7 @@ def embedVJ(genelist, maplist):
             ind[find] = 1
             VJ_array.append(ind)
         except ValueError:
-            print("Gene out of bound!" + gene)
+            print("Gene out of bound!" + str(gene))
             VJ_array.append(ind)
             next
     return np.asarray(VJ_array)

biopipen/scripts/tcr/TESSA_source/real_data.R

@@ -31,7 +31,7 @@ run_tessa <- function(
     # (optional) sample_id: a column vector of sample categories. If is_sampleCluster=TRUE, users must provide an additional
     #   column next to the cdr3 column.
     # (optional) fixed_b: a vector of pre-defined b. The vector must be numerical and has the length of TCR embeddings.
-    exp_data <- read.csv(exp_file, row.names=1, stringsAsFactors=F)
+    exp_data <- read.csv(exp_file, row.names=1, stringsAsFactors=F, check.names = FALSE)
     n <- ncol(exp_data)
     tmp <- apply(exp_data, 1, sd)
     # Run TSNE

{biopipen-0.17.7.dist-info → biopipen-0.18.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: biopipen
-Version: 0.17.7
+Version: 0.18.0
 Summary: Bioinformatics processes/pipelines that can be run from `pipen run`
 License: MIT
 Author: pwwang
@@ -13,10 +13,10 @@ Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Provides-Extra: runinfo
-Requires-Dist: datar[pandas] (>=0.13,<0.14)
-Requires-Dist: pipen (>=0.10,<0.11)
-Requires-Dist: pipen-board[report] (>=0.11,<0.12)
-Requires-Dist: pipen-cli-run (>=0.9,<0.10)
-Requires-Dist: pipen-filters (>=0.8,<0.9)
-Requires-Dist: pipen-runinfo (>=0.2,<0.3) ; extra == "runinfo"
-Requires-Dist: pipen-verbose (>=0.7,<0.8)
+Requires-Dist: datar[pandas] (>=0.15.2,<0.16.0)
+Requires-Dist: pipen (>=0.11,<0.12)
+Requires-Dist: pipen-board[report] (>=0.12,<0.13)
+Requires-Dist: pipen-cli-run (>=0.10,<0.11)
+Requires-Dist: pipen-filters (>=0.9,<0.10)
+Requires-Dist: pipen-runinfo (>=0.3,<0.4) ; extra == "runinfo"
+Requires-Dist: pipen-verbose (>=0.8,<0.9)