MinIONQC - segfault due to various issues

[steffcruve-github]

Hi,

Just came through this issue. I first thought it was a problem due to to nfs storage and try with a prior local copy of the file. Below is part of what is thrown by MinionQC (i.e. R). Interestingly, from a signle gridion run, some libraries got processed successfully and some others did not. Other thing to mention, I have been able to process successfully the "weird" files locally on my Mac.

I installed MinionQC via conda (bioconda repo) and have the version 1.4.1.

Could you help on that?

Best.

`(base) [Ztest]$ cp /nfs/SEQ_Nanopore/R-19-0167-6/L-19-1406-5/20191219_1649_GA20000_FAL80778_4d440b92/GXB02046_20191219_174922_FAL80778_gridion_sequencing_run_L-19-1406-5_sequencing_summary.txt sequencing_summary.txt && MinIONQC.R -i sequencing_summary.txt -o .
During startup - Warning messages:
1: Setting LC_CTYPE failed, using "C"
2: Setting LC_COLLATE failed, using "C"
3: Setting LC_TIME failed, using "C"
4: Setting LC_MESSAGES failed, using "C"
5: Setting LC_MONETARY failed, using "C"
6: Setting LC_PAPER failed, using "C"
7: Setting LC_MEASUREMENT failed, using "C"
INFO [2020-10-26 12:15:39] Loading input file: sequencing_summary.txt
INFO [2020-10-26 12:15:41] MinION flowcell detected
INFO [2020-10-26 12:15:51] .: creating output directory:./.
INFO [2020-10-26 12:15:51] .: summarising input file for flowcell
INFO [2020-10-26 12:15:53] .: plotting length histogram
INFO [2020-10-26 12:15:56] .: plotting mean Q score histogram
INFO [2020-10-26 12:15:58] .: plotting flowcell overview
INFO [2020-10-26 12:16:54] .: plotting flowcell yield over time
INFO [2020-10-26 12:17:10] .: plotting flowcell yield by read length
INFO [2020-10-26 12:17:25] .: plotting sequence length over time

*** caught segfault ***
address 0x7f21f504d040, cause 'memory not mapped'

*** caught segfault ***
address 0x7f21fa04f040, cause 'invalid permissions'

*** caught segfault ***
address 0x7f21f784e040, cause 'invalid permissions'

Traceback:
1: gam.fit3.post.proc(G$X, G$L, G$lsp0, G$S, G$off, object)

Traceback:
1: gam.fit3.post.proc(G$X, G$L, G$lsp0, G$S, G$off, object)
2: gam.outer(lsp, fscale = null.stuff$null.scale, family = G$family, control = control, criterion = criterion, method = method, optimizer = optimizer, scale = scale, gamma = gamma, G = G, start = start, ...)
3: estimate.gam(G, method, optimizer, control, in.out, scale, gamma, ...)
4: (function (formula, family = gaussian(), data = list(), weights = NULL, subset = NULL, na.action, offset = NULL, method = "GCV.Cp", optimizer = c("outer", "newton"), control = list(), scale = , select = FALSE, knots = NULL, sp = NULL, min.sp = NULL, H = NULL, gamma = 1, fit = TRUE, paraPen = NULL, G = NULL, in.out = NULL, drop.unused.levels = TRUE, drop.intercept = NULL, discrete = FALSE, ...) { control <- do.call("gam.control", control) if (is.null(G) && discrete) { cl <- match.call() cl[[1]] <- quote(bam) cl$fit = FALSE G <- eval(cl, parent.frame()) } if (is.null(G)) { gp <- interpret.gam(formula) cl <- match.call() mf <- match.call(expand.dots = FALSE) mf$formula <- gp$fake.formula mf$family <- mf$control <- mf$scale <- mf$knots <- mf$sp <- mf$min.sp <- mf$H <- mf$select <- mf$drop.intercept <- mf$gamma <- mf$method <- mf$fit <- mf$paraPen <- mf$G <- mf$optimizer <- mf$in.out <- mf$discrete <- mf$... <- NULL mf$drop.unused.levels <- drop.unused.levels mf[[1]] <- quote(stats::model.frame) pmf <- mf mf <- eval(mf, parent.frame()) if (nrow(mf) < 2) stop("Not enough (non-NA) data to do anything meaningful") terms <- attr(mf, "terms") vars <- all.vars1(gp$fake.formula[-2]) inp <- parse(text = paste("list(", paste(vars, collapse = ","), ")")) if (!is.list(data) && !is.data.frame(data)) data <- as.data.frame(data) dl <- eval(inp, data, parent.frame()) names(dl) <- vars var.summary <- variable.summary(gp$pf, dl, nrow(mf)) rm(dl) if (is.list(formula)) { environment(formula) <- environment(formula[[1]]) pterms <- list() tlab <- rep("", 0) for (i in 1:length(formula)) { pmf$formula <- gp[[i]]$pf pterms[[i]] <- attr(eval(pmf, parent.frame()), "terms") tlabi <- attr(pterms[[i]], "term.labels") if (i > 1 && length(tlabi) > 0) tlabi <- paste(tlabi, i - 1, sep = ".") tlab <- c(tlab, tlabi) } attr(pterms, "term.labels") <- tlab } else { pmf$formula <- gp$pf pmf <- eval(pmf, parent.frame()) pterms <- attr(pmf, "terms") } if (is.character(family)) family <- eval(parse(text = family)) if (is.function(family)) family <- family() if (is.null(family$family)) stop("family not recognized") if (family$family[1] == "gaussian" && family$link == "identity") am <- TRUE else am <- FALSE if (!control$keepData) rm(data) if (is.null(family$drop.intercept)) { lengthf <- if (is.list(formula)) length(formula) else 1 if (is.null(drop.intercept)) drop.intercept <- rep(FALSE, lengthf) else { drop.intercept <- rep(drop.intercept, length = lengthf) if (sum(drop.intercept)) family$drop.intercept <- drop.intercept } } else drop.intercept <- as.logical(family$drop.intercept) if (inherits(family, "general.family") && !is.null(family$presetup)) eval(family$presetup) gsname <- if (is.list(formula)) "gam.setup.list" else "gam.setup" G <- do.call(gsname, list(formula = gp, pterms = pterms, data = mf, knots = knots, sp = sp, min.sp = min.sp, H = H, absorb.cons = TRUE, sparse.cons = 0, select = select, idLinksBases = control$idLinksBases, scale.penalty = control$scalePenalty, paraPen = paraPen, drop.intercept = drop.intercept)) G$var.summary <- var.summary G$family <- family if ((is.list(formula) && (is.null(family$nlp) || family$nlp != gp$nlp)) || (!is.list(formula) && !is.null(family$npl) && (family$npl > 1))) stop("incorrect number of linear predictors for family") G$terms <- terms G$mf <- mf G$cl <- cl G$am <- am if (is.null(G$offset)) G$offset <- rep(0, G$n) G$min.edf <- G$nsdf if (G$m) for (i in 1:G$m) G$min.edf <- G$min.edf + G$smooth[[i]]$null.space.dim G$formula <- formula G$pred.formula <- gp$pred.formula environment(G$formula) <- environment(formula) } else { if (!is.null(sp) && any(sp >= 0)) { if (is.null(G$L)) G$L <- diag(length(G$sp)) if (length(sp) != ncol(G$L)) stop("length of sp must be number of free smoothing parameters in original model") ind <- sp >= 0 spind <- log(sp[ind]) spind[!is.finite(spind)] <- -30 G$lsp0 <- G$lsp0 + drop(G$L[, ind, drop = FALSE] %% spind) G$L <- G$L[, !ind, drop = FALSE] G$sp <- rep(-1, ncol(G$L)) } } if (!fit) { class(G) <- "gam.prefit" return(G) } if (ncol(G$X) > nrow(G$X)) stop("Model has more coefficients than data") G$conv.tol <- control$mgcv.tol G$max.half <- control$mgcv.half object <- estimate.gam(G, method, optimizer, control, in.out, scale, gamma, ...) if (!is.null(G$L)) { object$full.sp <- as.numeric(exp(G$L %% log(object$sp) + G$lsp0)) names(object$full.sp) <- names(G$lsp0) } names(object$sp) <- names(G$sp) object$paraPen <- G$pP object$formula <- G$formula if (is.list(object$formula)) attr(object$formula, "lpi") <- attr(G$X, "lpi") object$var.summary <- G$var.summary object$cmX <- G$cmX object$model <- G$mf object$na.action <- attr(G$mf, "na.action") object$control <- control object$terms <- G$terms object$pred.formula <- G$pred.formula attr(object$pred.formula, "full") <- reformulate(all.vars(object$terms)) object$pterms <- G$pterms object$assign <- G$assign object$contrasts <- G$contrasts object$xlevels <- G$xlevels object$offset <- G$offset if (!is.null(G$Xcentre)) object$Xcentre <- G$Xcentre if (control$keepData) object$data <- data object$df.residual <- nrow(G$X) - sum(object$edf) object$min.edf <- G$min.edf if (G$am && !(method %in% c("REML", "ML", "P-ML", "P-REML"))) object$optimizer <- "magic" else object$optimizer <- optimizer object$call <- G$cl class(object) <- c("gam", "glm", "lm") if (is.null(object$deviance)) object$deviance <- sum(residuals(object, "deviance")^2) names(object$gcv.ubre) <- method environment(object$formula) <- environment(object$pred.formula) <- environment(object$terms) <- environment(object$pterms) <- .GlobalEnv if (!is.null(object$model)) environment(attr(object$model, "terms")) <- .GlobalEnv if (!is.null(attr(object$pred.formula, "full"))) environment(attr(object$pred.formula, "full")) <- .GlobalEnv object})(formula, data = data, weights = weight, method = "REML")
5: do.call(method, c(base.args, method.args))
6: f(...)
7: self$compute_group(data = group, scales = scales, ...)
8: FUN(X[[i]], ...)
9: lapply(groups, function(group) { self$compute_group(data = group, scales = scales, ...)})
10: f(..., self = self)
11: (function (...) f(..., self = self))(data = data, scales = scales, na.rm = FALSE, se = TRUE, flipped_aes = FALSE, method = function (formula, family = gaussian(), data = list(), weights = NULL, subset = NULL, na.action, offset = NULL, method = "GCV.Cp", optimizer = c("outer", "newton"), control = list(), scale = 0, select = FALSE, knots = NULL, sp = NULL, min.sp = NULL, H = NULL, gamma = 1, fit = TRUE, paraPen = NULL, G = NULL, in.out = NULL, drop.unused.levels = TRUE, drop.intercept = NULL, discrete = FALSE, ...) { control <- do.call("gam.control", control) if (is.null(G) && discrete) { cl <- match.call() cl[[1]] <- quote(bam) cl$fit = FALSE G <- eval(cl, parent.frame()) } if (is.null(G)) { gp <- interpret.gam(formula) cl <- match.call() mf <- match.call(expand.dots = FALSE) mf$formula <- gp$fake.formula mf$family <- mf$control <- mf$scale <- mf$knots <- mf$sp <- mf$min.sp <- mf$H <- mf$select <- mf$drop.intercept <- mf$gamma <- mf$method <- mf$fit <- mf$paraPen <- mf$G <- mf$optimizer <- mf$in.out <- mf$discrete <- mf$... <- NULL mf$drop.unused.levels <- drop.unused.levels mf[[1]] <- quote(stats::model.frame) pmf <- mf mf <- eval(mf, parent.frame()) if (nrow(mf) < 2) stop("Not enough (non-NA) data to do anything meaningful") terms <- attr(mf, "terms") vars <- all.vars1(gp$fake.formula[-2]) inp <- parse(text = paste("list(", paste(vars, collapse = ","), ")")) if (!is.list(data) && !is.data.frame(data)) data <- as.data.frame(data) dl <- eval(inp, data, parent.frame()) names(dl) <- vars var.summary <- variable.summary(gp$pf, dl, nrow(mf)) rm(dl) if (is.list(formula)) { environment(formula) <- environment(formula[[1]]) pterms <- list() tlab <- rep("", 0) for (i in 1:length(formula)) { pmf$formula <- gp[[i]]$pf pterms[[i]] <- attr(eval(pmf, parent.frame()), "terms") tlabi <- attr(pterms[[i]], "term.labels") if (i > 1 && length(tlabi) > 0) tlabi <- paste(tlabi, i - 1, sep = ".") tlab <- c(tlab, tlabi) } attr(pterms, "term.labels") <- tlab } else { pmf$formula <- gp$pf pmf <- eval(pmf, parent.frame()) pterms <- attr(pmf, "terms") } if (is.character(family)) family <- eval(parse(text = family)) if (is.function(family)) family <- family() if (is.null(family$family)) stop("family not recognized") if (family$family[1] == "gaussian" && family$link == "identity") am <- TRUE else am <- FALSE if (!control$keepData) rm(data) if (is.null(family$drop.intercept)) { lengthf <- if (is.list(formula)) length(formula) else 1 if (is.null(drop.intercept)) drop.intercept <- rep(FALSE, lengthf) else { drop.intercept <- rep(drop.intercept, length = lengthf) if (sum(drop.intercept)) family$drop.intercept <- drop.intercept } } else drop.intercept <- as.logical(family$drop.intercept) if (inherits(family, "general.family") && !is.null(family$presetup)) eval(family$presetup) gsname <- if (is.list(formula)) "gam.setup.list" else "gam.setup" G <- do.call(gsname, list(formula = gp, pterms = pterms, data = mf, knots = knots, sp = sp, min.sp = min.sp, H = H, absorb.cons = TRUE, sparse.cons = 0, select = select, idLinksBases = control$idLinksBases, scale.penalty = control$scalePenalty, paraPen = paraPen, drop.intercept = drop.intercept)) G$var.summary <- var.summary G$family <- family if ((is.list(formula) && (is.null(family$nlp) || family$nlp != gp$nlp)) || (!is.list(formula) && !is.null(family$npl) && (family$npl > 1))) stop("incorrect number of linear predictors for family") G$terms <- terms G$mf <- mf G$cl <- cl G$am <- am if (is.null(G$offset)) G$offset <- rep(0, G$n) G$min.edf <- G$nsdf if (G$m) for (i in 1:G$m) G$min.edf <- G$min.edf + G$smooth[[i]]$null.space.dim G$formula <- formula G$pred.formula <- gp$pred.formula environment(G$formula) <- environment(formula) } else { if (!is.null(sp) && any(sp >= 0)) { if (is.null(G$L)) G$L <- diag(length(G$sp)) if (length(sp) != ncol(G$L)) stop("length of sp must be number of free smoothing parameters in original model") ind <- sp >= 0 spind <- log(sp[ind]) spind[!is.finite(spind)] <- -30 G$lsp0 <- G$lsp0 + drop(G$L[, ind, drop = FALSE] %% spind) G$L <- G$L[, !ind, drop = FALSE] G$sp <- rep(-1, ncol(G$L)) } } if (!fit) { class(G) <- "gam.prefit" return(G) } if (ncol(G$X) > nrow(G$X)) stop("Model has more coefficients than data") G$conv.tol <- control$mgcv.tol G$max.half <- control$mgcv.half object <- estimate.gam(G, method, optimizer, control, in.out, scale, gamma, ...) if (!is.null(G$L)) { object$full.sp <- as.numeric(exp(G$L %% log(object$sp) + G$lsp0)) names(object$full.sp) <- names(G$lsp0) } names(object$sp) <- names(G$sp) object$paraPen <- G$pP object$formula <- G$formula if (is.list(object$formula)) attr(object$formula, "lpi") <- attr(G$X, "lpi") object$var.summary <- G$var.summary object$cmX <- G$cmX object$model <- G$mf object$na.action <- attr(G$mf, "na.action") object$control <- control object$terms <- G$terms object$pred.formula <- G$pred.formula attr(object$pred.formula, "full") <- reformulate(all.vars(object$terms)) object$pterms <- G$pterms object$assign <- G$assign object$contrasts <- G$contrasts object$xlevels <- G$xlevels object$offset <- G$offset if (!is.null(G$Xcentre)) object$Xcentre <- G$Xcentre if (control$keepData) object$data <- data object$df.residual <- nrow(G$X) - sum(object$edf) object$min.edf <- G$min.edf if (G$am && !(method %in% c("REML", "ML", "P-ML", "P-REML"))) object$optimizer <- "magic" else object$optimizer <- optimizer object$call <- G$cl class(object) <- c("gam", "glm", "lm") if (is.null(object$deviance)) object$deviance <- sum(residuals(object, "deviance")^2) names(object$gcv.ubre) <- method environment(object$formula) <- environment(object$pred.formula) <- environment(object$terms) <- environment(object$pterms) <- .GlobalEnv if (!is.null(object$model)) environment(attr(object$model, "terms")) <- .GlobalEnv if (!is.null(attr(object$pred.formula, "full"))) environment(attr(object$pred.formula, "full")) <- .GlobalEnv object }, formula = y ~ s(x, bs = "cs"))
12: do.call(self$compute_panel, args)
13: doTryCatch(return(expr), name, parentenv, handler)
14: tryCatchOne(expr, names, parentenv, handlers[[1L]])
15: tryCatchList(expr, classes, parentenv, handlers)
16: tryCatch(do.call(self$compute_panel, args), error = function(e) { warn(glue("Computation failed in {snake_class(self)}():\n{e$message}")) new_data_frame()})
17: fun(x, ...)
18: apply_fun(df)
19: dapply(data, "PANEL", function(data) { scales <- layout$get_scales(data$PANEL[1]) tryCatch(do.call(self$compute_panel, args), error = function(e) { warn(glue("Computation failed in {snake_class(self)}():\n{e$message}")) new_data_frame() })})
20: f(..., self = self)
21: self$stat$compute_layer(data, params, layout)
22: f(..., self = self)
23: l$compute_statistic(d, layout)
24: f(l = layers[[i]], d = data[[i]])
25: by_layer(function(l, d) l$compute_statistic(d, layout))
26: ggplot_build.ggplot(x)
27: ggplot_build(x)
28: print.ggplot(x)
29: print(x)
30: grid.draw.ggplot(plot)
31: grid.draw(plot)
32: ggsave(filename = file.path(output.dir, "length_by_hour.png"), width = p1m * 960/75, height = p1m * 480/75, plot = p7)
33: withCallingHandlers(expr, message = function(c) invokeRestart("muffleMessage"))
34: suppressMessages(ggsave(filename = file.path(output.dir, "length_by_hour.png"), width = p1m * 960/75, height = p1m * 480/75, plot = p7))
35: single.flowcell(input.file, output.dir, q)
An irrecoverable exception occurred. R is aborting now ...`