biopipen 0.28.0__py3-none-any.whl → 0.29.0__py3-none-any.whl

biopipen/scripts/snp/PlinkIBD.R

@@ -0,0 +1,197 @@
+source("{{biopipen_dir}}/utils/misc.R")
+source("{{biopipen_dir}}/utils/plot.R")
+suppressPackageStartupMessages({
+    library(dplyr)
+    library(tidyr)
+    library(tibble)
+})
+
+indir    <- {{in.indir | r}}
+outdir   <- {{out.outdir | r}}
+plink    <- {{envs.plink | r}}
+indep    <- {{envs.indep | r}}
+highld   <- {{envs.highld | r}}
+devpars  <- {{envs.devpars | r}}
+pihat    <- {{envs.pihat | r}}
+samid    <- {{envs.samid | r}}
+annofile <- {{envs.anno | r}}
+doplot   <- {{envs.plot | r}}
+seed     <- {{envs.seed | r}}
+ncores   <- {{envs.ncores | r}}
+
+bedfile <- Sys.glob(file.path(indir, '*.bed'))
+if (length(bedfile) == 0)
+    stop("No bed files found in the input directory.")
+if (length(bedfile) > 1) {
+    log_warn("Multiple bed files found in the input directory. Using the first one.")
+    bedfile <- bedfile[1]
+}
+input  <- tools::file_path_sans_ext(bedfile)
+output <- file.path(outdir, basename(input))
+
+cmd <- c(
+    plink,
+    "--threads", ncores,
+    "--bfile", input,
+    "--indep-pairwise", indep,
+	# One should be mindful of running this with < 50 samples
+	# "--bad-ld",
+    "--out", output
+)
+if (!is.null(highld) && !isFALSE(highld)) {
+    cmd <- c(cmd, "--range", "--exclude", highld)
+}
+run_command(cmd, fg = TRUE)
+
+prunein <- paste0(output, '.prune.in')
+cmd <- c(
+    plink,
+    "--threads", ncores,
+    "--bfile", input,
+    "--extract", prunein,
+    "--genome",
+    "--out", output
+)
+run_command(cmd, fg = TRUE)
+
+genome <- read.table(
+    paste0(output, '.genome'),
+    row.names = NULL,
+    header = TRUE,
+    check.names = FALSE
+)
+# "unmelt" it
+# FID1 IID1 FID2 IID2 RT EZ Z0     Z1     Z2     PI_HAT PHE DST      PPC    RATIO
+# s1   s1   s2   s2   UN NA 1.0000 0.0000 0.0000 0.0000 -1  0.866584 0.0000 0.9194
+# s1   s1   s2   s2   UN NA 0.4846 0.3724 0.1431 0.3293 -1  0.913945 0.7236 2.0375
+# s1   s1   s3   s3   UN NA 1.0000 0.0000 0.0000 0.0000 -1  0.867186 0.0000 1.0791
+genome$SAMPLE1 <- paste(genome$FID1, genome$IID1, sep = "\t")
+genome$SAMPLE2 <- paste(genome$FID2, genome$IID2, sep = "\t")
+
+
+# get all samples
+samples <- unique(c(genome$SAMPLE1, genome$SAMPLE2))
+# make paired into a distance-like matrix
+similarity <- genome %>%
+    select(SAMPLE1, SAMPLE2, PI_HAT) %>%
+    pivot_wider(names_from = SAMPLE2, values_from = PI_HAT, values_fill = NA) %>%
+    as.data.frame() %>%
+    column_to_rownames("SAMPLE1")
+rm(genome)
+# get the rownames back
+samids <- rownames(similarity)
+# get samples that didn't involved
+missedrow <- setdiff(samples, rownames(similarity))
+missedcol <- setdiff(samples, colnames(similarity))
+similarity[missedrow, ] <- NA
+similarity[, missedcol] <- NA
+# order the matrix
+similarity <- similarity[samples, samples, drop = FALSE]
+# transpose the matrix to get the symmetric values
+sim2 <- t(similarity)
+isna <- is.na(similarity)
+# fill the na's with their symmetric values
+similarity[isna] <- sim2[isna]
+rm(sim2)
+# still missing: keep them
+similarity[is.na(similarity)] <- 0
+# get the marks (samples that fail the pihat cutoff)
+nsams <- length(samples)
+fails <- which(similarity > pihat)
+marks <- data.frame(x = (fails - 1)%%nsams + 1, y = ceiling(fails/nsams))
+diag(similarity) <- 1
+
+failflags <- rep(F, nrow(marks))
+freqs <- as.data.frame(table(factor(as.matrix(marks))))
+freqs <- freqs[order(freqs$Freq, decreasing = T), 'Var1', drop = T]
+ibd.fail <- c()
+while (sum(failflags) < nrow(marks)) {
+	samidx <- freqs[1]
+	ibd.fail <- c(ibd.fail, samples[samidx])
+	freqs <- freqs[-1]
+	sapply(1:nrow(marks), function(i) {
+		if (samidx %in% marks[i,])
+			failflags[i] <<- TRUE
+	})
+}
+
+ibd_fail_file <- paste0(output, '.ibd.fail')
+writeLines(ibd.fail, ibd_fail_file)
+cmd <- c(
+    plink,
+    "--threads", ncores,
+    "--bfile", input,
+    "--remove", ibd_fail_file,
+	"--make-bed",
+    "--out", output
+)
+run_command(cmd, fg = TRUE)
+
+if (doplot) {
+	set.seed(seed)
+	library(ComplexHeatmap)
+	fontsize8 <- gpar(fontsize = 8)
+	fontsize9 <- gpar(fontsize = 9)
+	ht_opt$heatmap_row_names_gp <- fontsize8
+	ht_opt$heatmap_column_names_gp <- fontsize8
+	ht_opt$legend_title_gp <- fontsize9
+	ht_opt$legend_labels_gp <- fontsize8
+	ht_opt$simple_anno_size <- unit(3, "mm")
+
+	samids <- sapply(samples, function(sid) {
+		fidiid <- unlist(strsplit(sid, "\t", fixed = TRUE))
+		gsub(
+            "{fid}",
+            fidiid[1],
+            gsub("{iid}", fidiid[2], samid, fixed = TRUE),
+            fixed = TRUE
+        )
+	})
+	rownames(similarity) <- samids
+	colnames(similarity) <- samids
+
+	annos <- list()
+	if (!is.null(annofile) && !isFALSE(annofile)) {
+		options(stringsAsFactors = TRUE)
+		andata <- read.table(annofile, header = TRUE, row.names = 1, sep = "\t", check.names = FALSE)
+		andata <- andata[samids, , drop = FALSE]
+		for (anname in colnames(andata)) {
+			annos[[anname]] <- as.matrix(andata[, anname])
+		}
+		annos$annotation_name_gp <- fontsize8
+		annos <- do.call(HeatmapAnnotation, annos)
+	}
+
+	args <- list(
+		name = "PI_HAT",
+		cell_fun = function(j, i, x, y, width, height, fill) {
+			if (similarity[i, j] > pihat && i != j)
+				grid.points(x, y, pch = 4, size = unit(.5, "char"))
+		},
+		#heatmap_legend_param = list(
+		#	title_gp  = fontsize9,
+		#	labels_gp = fontsize8
+		#),
+		clustering_distance_rows = function(m) as.dist(1-m),
+		clustering_distance_columns = function(m) as.dist(1-m),
+		top_annotation = if (length(annos) == 0) NULL else annos
+	)
+
+	plotHeatmap(
+		similarity,
+		outfile = paste0(output, '.ibd.png'),
+		args = args,
+		draw = list(
+			annotation_legend_list = list(
+				Legend(
+					labels = paste(">", pihat),
+					title = "",
+					type = "points",
+					pch = 4,
+					title_gp = fontsize9,
+					labels_gp = fontsize8)),
+			merge_legend = TRUE
+        ),
+		devpars = devpars
+    )
+}

biopipen/scripts/snp/PlinkUpdateName.py

@@ -0,0 +1,124 @@
+from pathlib import Path
+from biopipen.utils.misc import run_command, dict_to_cli_args, logger
+
+indir = {{in.indir | repr}}  # pyright: ignore # noqa: #999
+namefile = {{in.namefile | repr}}  # pyright: ignore
+outdir = {{out.outdir | repr}}  # pyright: ignore
+plink = {{envs.plink | repr}}  # pyright: ignore
+bcftools = {{envs.bcftools | repr}}  # pyright: ignore
+ncores = {{envs.ncores | repr}}  # pyright: ignore
+match_alt = {{envs.match_alt | repr}}  # pyright: ignore
+
+bedfile = list(Path(indir).glob("*.bed"))
+if len(bedfile) == 0:
+    raise FileNotFoundError(f"No .bed file found in `in.indir`")
+elif len(bedfile) > 1:
+    logger.warning(f"Multiple .bed files found in `in.indir`, using the first one.")
+
+bedfile = bedfile[0]
+input = bedfile.with_suffix("")
+output = Path(outdir) / bedfile.stem
+
+if namefile.endswith(".vcf") or namefile.endswith(".vcf.gz"):
+    logger.info("VCF file received, extracting names")
+    def alt_matched(bim_alt, vcf_alt, match_alt):
+        if match_alt == "none":
+            return True
+        if match_alt == "exact":
+            return bim_alt == vcf_alt
+
+        bim_alts = bim_alt.split(",")
+        vcf_alts = vcf_alt.split(",")
+        if match_alt == "all":
+            return set(bim_alts) == set(vcf_alts)
+        if match_alt == "any":
+            return bool(set(bim_alts) & set(vcf_alts))
+        if match_alt == "first_included":
+            return bim_alts[0] in vcf_alts
+        if match_alt == "first":
+            return bim_alts[0] == vcf_alts[0]
+
+        raise ValueError(f"Unknown match_alt: {match_alt}")
+
+    def readline(f):
+        line = f.readline().strip()
+        return line.split("\t") if line else None
+
+    namefile_tmp = Path(outdir) / "_namefile_from_vcf.txt"
+    infofile = Path(outdir) / "_information_from_vcf_unsorted.txt"
+    sorted_infofile = Path(outdir) / "_information_from_vcf_sorted.txt"
+    sorted_bim = Path(outdir) / "_sorted_bim.txt"
+    bt_cmd = [
+        bcftools, "query",
+        "-f", "%CHROM\\t%ID\\t0\\t%POS\\t%ALT\\t%REF\\n",
+        "-o", infofile,
+        namefile,
+    ]
+    ## infofile
+    # 1	rs10492	0	10492	T	C
+    logger.info("- Extracting information from VCF file ...")
+    run_command(bt_cmd, fg=True)
+    # sort infofile
+    logger.info("- Sorting the information from VCF file ...")
+    run_command(
+        [
+            "sort",
+            "-k1,1", "-k4,4n", "-k6,6",
+            infofile,
+            "--parallel", ncores,
+            "-o", sorted_infofile
+        ],
+        env={"LC_ALL": "C"},
+        fg=True,
+    )
+
+    ## .bim file
+    # 1	1_10492	0	10492	T	C
+    # sort .bim file
+    logger.info("- Sorting the .bim file ...")
+    run_command(
+        [
+            "sort",
+            "-k1,1", "-k4,4n", "-k6,6",
+            input.with_suffix(".bim"),
+            "--parallel", ncores,
+            "-o", sorted_bim
+        ],
+        env={"LC_ALL": "C"},
+        fg=True,
+    )
+    # query namefile for records in sorted bim file
+    logger.info("- Matching and generating the name file ...")
+    with sorted_bim.open() as fbim, sorted_infofile.open() as finfo, namefile_tmp.open("w") as fout:  # noqa: E501
+        bim = readline(fbim)
+        info = readline(finfo)
+        while bim and info:
+            if (
+                bim[0] == info[0]
+                and bim[3] == info[3]
+                and bim[5] == info[5]
+                and alt_matched(bim[4], info[4], match_alt)
+            ):
+                fout.write(f"{bim[1]}\t{info[1]}\n")
+                bim = readline(fbim)
+                info = readline(finfo)
+            elif (
+                bim[0] < info[0]
+                or (bim[0] == info[0] and bim[3] < info[3])
+                or (bim[0] == info[0] and bim[3] == info[3] and bim[5] < info[5])
+            ):
+                bim = readline(fbim)
+            else:
+                info = readline(finfo)
+
+    namefile = namefile_tmp
+
+args = {
+    "": plink,
+    "bfile": input,
+    "out": output,
+    "make_bed": True,
+    "update_name": namefile,
+}
+
+run_command(dict_to_cli_args(args, dashify=True), fg=True)

biopipen/scripts/stats/MetaPvalue.R

@@ -11,6 +11,7 @@ id_exprs <- {{envs.id_exprs | r}}
 pval_cols <- {{envs.pval_cols | r}}
 method <- {{envs.method | r}}
 na <- {{envs.na | r}}
+keep_single <- {{envs.keep_single | r}}
 padj <- {{envs.padj | r}}
 
 if (method == "fisher") { method = "sumlog" }
@@ -102,7 +103,7 @@ if (length(infiles) == 1 && padj == "none") {
         if (length(ps) == 0) {
             metaps <- c(metaps, NA)
             ns <- c(ns, NA)
-        } else if (length(ps) == 1) {
+        } else if (length(ps) == 1 && keep_single) {
             metaps <- c(metaps, ps)
             ns <- c(ns, 1)
         } else {

biopipen/scripts/stats/MetaPvalue1.R

@@ -0,0 +1,70 @@
+source("{{biopipen_dir}}/utils/misc.R")
+
+library(metap)
+library(rlang)
+library(dplyr)
+
+infile <- {{in.infile | r}}
+outfile <- {{out.outfile | r}}
+id_cols <- {{envs.id_cols | r}}
+pval_col <- {{envs.pval_col | r}}
+method <- {{envs.method | r}}
+na <- {{envs.na | r}}
+keep_single <- {{envs.keep_single | r}}
+padj <- {{envs.padj | r}}
+
+if (method == "fisher") { method = "sumlog" }
+
+# Check pval_cols
+if (is.null(pval_col)) { stop("Must provide envs.pval_col") }
+
+# Check id_cols
+if (is.null(id_cols)) { stop("Must provide envs.id_cols") }
+if (length(id_cols) == 1) {
+    id_cols <- trimws(strsplit(id_cols, ",")[[1]])
+}
+
+log_info("Reading input and performing meta-analysis ...")
+outdata <- read.table(
+        infile, header = TRUE, sep = "\t", row.names = NULL, check.names = FALSE
+    ) %>%
+    group_by(!!!syms(id_cols)) %>%
+    summarise(
+        N = n(),
+        .pvals = list(!!sym(pval_col)),
+        .groups = "drop"
+    )
+
+metaps <- c()
+ns <- c()
+for (ps in outdata$.pvals) {
+    if (na == -1) {
+        ps <- ps[!is.na(ps)]
+    } else {
+        ps[is.na(ps)] <- na
+    }
+
+    if (length(ps) == 0) {
+        metaps <- c(metaps, NA)
+        ns <- c(ns, NA)
+    } else if (length(ps) == 1 && keep_single) {
+        metaps <- c(metaps, ps)
+        ns <- c(ns, 1)
+    } else {
+        metaps <- c(metaps, do.call(method, list(ps))$p)
+        ns <- c(ns, length(ps))
+    }
+}
+outdata$MetaPval <- metaps
+outdata$N <- ns
+outdata$.pvals <- NULL
+outdata <- outdata %>% arrange(MetaPval)
+
+if (padj != "none") {
+    log_info("Calculating adjusted p-values ...")
+    outdata$MetaPadj <- p.adjust(outdata$MetaPval, method = padj)
+
+}
+
+log_info("Writing output ...")
+write.table(outdata, outfile, quote = FALSE, sep = "\t", row.names = FALSE)

biopipen/scripts/tcr/TCRClusterStats.R

@@ -130,13 +130,6 @@ shared_clusters = function(name) {
         row.names=TRUE, col.names=TRUE, quote=FALSE, sep="\t"
     )
 
-    if (is.null(case$heatmap_meta) || length(case$heatmap_meta) == 0) {
-        anno = NULL
-    } else {
-        anno = as.list(immdata$meta[, case$heatmap_meta, drop=FALSE])
-        anno = do_call(ComplexHeatmap::HeatmapAnnotation, anno)
-    }
-
     if (!is.null(case$sample_order) && length(case$sample_order) > 0) {
         if (length(case$sample_order) == 1) {
             case$sample_order = trimws(strsplit(case$sample_order, ",")[[1]])
@@ -148,6 +141,18 @@ shared_clusters = function(name) {
         plotdata = plotdata[, case$sample_order, drop=FALSE]
     }
 
+    if (is.null(case$heatmap_meta) || length(case$heatmap_meta) == 0) {
+        anno = NULL
+    } else {
+        anno = as.list(
+            immdata$meta[
+                match(colnames(plotdata), immdata$meta$Sample),
+                case$heatmap_meta,
+                drop=FALSE
+            ])
+        anno = do_call(ComplexHeatmap::HeatmapAnnotation, anno)
+    }
+
     cluster_rows = case$cluster_rows && nrow(plotdata) > 2
     col_samples = colnames(plotdata)
     if (!cluster_rows) {

biopipen/scripts/vcf/BcftoolsAnnotate.py

@@ -0,0 +1,91 @@
+from os import path
+from contextlib import suppress
+from pathlib import PosixPath  # noqa: F401
+
+from biopipen.utils.reference import tabix_index
+from biopipen.utils.misc import logger
+from biopipen.scripts.vcf.bcftools_utils import run_bcftools
+
+infile = {{in.infile | repr}}  # pyright: ignore # noqa: E999
+annfile = {{in.annfile | repr}}  # pyright: ignore
+outfile = {{out.outfile | repr}}  # pyright: ignore
+joboutdir = {{job.outdir | repr}}  # pyright: ignore
+envs = {{envs | dict | repr}}  # pyright: ignore
+
+bcftools = envs.pop("bcftools")
+tabix = envs.pop("tabix")
+ncores = envs.pop("ncores")
+columns = envs.pop("columns")
+remove = envs.pop("remove")
+header = envs.pop("header")
+gz = envs.pop("gz")
+index = envs.pop("index")
+
+if isinstance(columns, list):
+    columns = ",".join(columns)
+
+if "c" in envs:
+    logger.warning("Ignoring envs\[c], use envs\[columns] instead.")
+    del envs["c"]
+
+if isinstance(remove, list):
+    remove = ",".join(remove)
+
+if "x" in envs:
+    logger.warning("Ignoring envs\[x], use envs\[remove] instead.")
+    del envs["x"]
+
+envs_has_annfile = "a" in envs or "annotations" in envs
+headerfile = path.join(joboutdir, "header.txt")
+if header:
+    with open(headerfile, "w") as fh:
+        fh.writelines(header)
+
+if annfile and envs_has_annfile:
+    logger.warning(
+        "Ignoring envs\[a/annotations] because in.annfile is provided."
+    )
+    with suppress(KeyError):
+        del envs["a"]
+    with suppress(KeyError):
+        del envs["annotations"]
+elif not annfile and envs_has_annfile:
+    annfile = envs.pop("annotations", None) or envs.pop("a", None)
+
+
+if index and not gz:
+    logger.warning("Forcing envs.gz to True because envs.index is True.")
+    gz = True
+
+envs[""] = [bcftools, "annotate"]
+envs["o"] = outfile
+envs["threads"] = ncores
+
+if "O" not in envs and "output-type" not in envs and "output_type" not in envs:
+    envs["O"] = "z" if gz else "v"
+
+if columns:
+    envs["columns"] = columns
+    if not annfile:
+        raise ValueError(
+            "envs.columns specified but no in.annfile/envs.annfile provided."
+        )
+    envs["_"] = tabix_index(infile, "vcf", tabix=tabix)
+
+if remove:
+    envs["remove"] = remove
+    # no need to index it
+    envs["_"] = infile
+
+if "columns" not in envs and "remove" not in envs:
+    logger.warning(
+        "No columns/remove specified, no columns will be carried over or removed."
+    )
+
+if annfile:
+    envs["annotations"] = tabix_index(annfile, "vcf", tabix=tabix)
+
+if header:
+    envs["header_lines"] = headerfile
+
+run_bcftools(envs, bcftools=bcftools, index=index, tabix=tabix)

biopipen/scripts/vcf/BcftoolsFilter.py

@@ -0,0 +1,90 @@
+from pathlib import Path, PosixPath  # noqa: F401
+
+from biopipen.utils.misc import logger
+from biopipen.scripts.vcf.bcftools_utils import run_bcftools
+
+infile = {{in.infile | repr}}  # pyright: ignore # noqa: #999
+outfile = {{out.outfile | repr}}  # pyright: ignore
+outdir = Path(outfile).parent
+
+envs = {{envs | dict | repr}}  # pyright: ignore
+bcftools = envs.pop("bcftools")
+tabix = envs.pop("tabix")
+keep = envs.pop("keep")
+ncores = envs.pop("ncores")
+includes = envs.pop("includes")
+excludes = envs.pop("excludes")
+gz = envs.pop("gz")
+index = envs.pop("index")
+
+# a.vcf.gz -> a
+# a.vcf -> a
+stem = Path(infile).stem
+if stem.endswith(".vcf"):
+    stem = stem[:-4]
+# .vcf.gz
+# .gz
+ext = ".vcf.gz" if index or gz else '.vcf'
+
+
+def normalize_expr(expr, flag, prev_n_filters=0):
+    out = {}
+    if not expr:
+        return out
+    if isinstance(expr, list):
+        for ex in expr:
+            out[f"FILTER_{flag.upper()}_{len(out) + 1 + prev_n_filters}"] = (ex, flag)
+    elif isinstance(expr, dict):
+        for name, ex in expr.items():
+            out[name] = (ex, flag)
+    else: # str
+        out[f"FILTER_{flag.upper()}_{len(out) + 1 + prev_n_filters}"] = (expr, flag)
+    return out
+
+
+def handle_filter(vcf, fname, filt, flag, final):
+    logger.info("- Handling filter %s: %s ...", fname, filt)
+
+    arguments = envs.copy()
+    arguments[flag] = filt
+    arguments["_"] = vcf
+    arguments["o"] = outfile if final else outdir / f"{stem}.{fname}{ext}"
+    if keep:
+        arguments["s"] = fname
+
+    run_bcftools(arguments, bcftools=bcftools, index=index and final, tabix=tabix)
+
+    if final:
+        flagfile = outdir.joinpath(f"{stem}.{fname}{ext}")
+        if flagfile.is_symlink():
+            flagfile.unlink()
+        outdir.joinpath(f"{stem}.{fname}{ext}").symlink_to(outfile)
+
+    return arguments["o"]
+
+
+includes = normalize_expr(includes, "include")
+excludes = normalize_expr(excludes, "exclude", len(includes))
+includes.update(excludes)
+
+if index and not gz:
+    logger.warning("Forcing envs.gz to True because envs.index is True.")
+    gz = True
+
+envs[""] = [bcftools, "filter"]
+envs["_"] = infile
+envs["o"] = outfile
+envs["threads"] = ncores
+
+if "O" not in envs and "output-type" not in envs and "output_type" not in envs:
+    envs["O"] = "z" if gz else "v"
+
+if keep:
+    envs["soft_filter"] = "+"
+
+if "m" not in envs and "mode" not in envs:
+    envs["m"] = "+"
+
+# bcftools can be only done once at one filter
+for i, (fname, (filt, flag)) in enumerate(includes.items()):
+    infile = handle_filter(infile, fname, filt, flag, i == len(includes) - 1)