The microdialysis-derived lactate-pyruvate gradient indicates anaerobic activity after brain injury.
10 blocks of code here have a permanent identifier (oscr:b:…), listed with the matches below: citable for good, with a page of its own. Select a match and press Cite to copy it. What a block identifier is.
The 15 matches
- [1] § Results › Substrate delivery and LPR fidelity ↔ 8_substrate_lpr_fidelity.R, lines 520 to 572 · score 0.87 · LPR fidelity, nested random intercepts, median glucose, LPR perfectly tracked, Patient random intercepts, hyperbolic LPR glucose · block
oscr:b:VPT211RGXEJMW-J· Report a mismatch - [2] § Results › Outcome association ↔ 9_outcome.R, lines 1 to 81 · score 0.83 · LME models, GOSE categories, segment median, Linear contrasts, 5–6, 7–8 · block
oscr:b:E2PKTJR09HRHV-I· Report a mismatch - [3] § Results › Linear parameters and the hyperbolic consequence › Characterisation ↔ 5_linear_LP_models.R, lines 1073 to 1131 · score 0.76 · Combined overlay comparing, Negative intercepts produce, Positive intercepts produce, cohort median trend, Dashed horizontal, LPR hyperbolae · block
oscr:b:ATX763TJK50Q5-F· Report a mismatch - [4] § Results › Linear parameters and the hyperbolic consequence › Within-segment LPR error ↔ 6_hyperbolic_LPR_error.R, lines 1137 to 1214 · score 0.74 · relative LPR error, pyruvate concentration, maximum pyruvate, zero intercept, minimum pyruvate, LPR range · block
oscr:b:9PWVBB3SNGZYS-D· Report a mismatch - [5] § Materials and methods › In vitro model ↔ 10_in_vitro.R, lines 1 to 60 · score 0.73 · electron transport chain, mitochondrial dysfunction, vitro, inhibitor, rotenone, model · block
oscr:b:X1HPPGJYHN9KV-K· Report a mismatch - [6] § Results › Substrate delivery and LPR fidelity ↔ 8_substrate_lpr_fidelity.R, lines 61 to 118 · score 0.70 · higher glucose, median glucose, anaerobic activity, median LPR, physiological, substrate · block
oscr:b:YAEMJKBV0F950-A· Report a mismatch - [7] § Results › Independent Uppsala cohort ↔ 8_substrate_lpr_fidelity.R, lines 520 to 572 · score 0.68 · residual SD, median glucose, LPR perfectly tracked, hyperbolic LPR glucose, median LPR, substrate · block
oscr:b:VPT211RGXEJMW-J· Report a mismatch - [8] § Results › Explaining metabolic polarisation ↔ 7_never_always_LPR25.R, lines 1 to 78 · score 0.68 · hyperbolic LPR pyruvate, higher intercepts, linear parameters, S3, threshold, pyruvate relationships · block
oscr:b:JMERCWB0991RX-S· Report a mismatch - [9] § Results › In vitro validation ↔ 10_in_vitro.R, lines 228 to 263 · score 0.67 · vitro lactate pyruvate, OLS regression fits, 1–24 h, rotenone treated, ANCOVA, R2 · block
oscr:b:5M8W8P3KF63SX-8· Report a mismatch - [10] § Materials and methods › Data integration and pre-processing ↔ 4_segmentation_stats.R, lines 1 to 72 · score 0.60 · pre processed, Segmentation pipeline, S4, S1, patient, linear · block
oscr:b:K2QQW4Q91XT22-6· Report a mismatch - [11] § Results › In vitro validation ↔ 10_in_vitro.R, lines 1 to 60 · score 0.59 · mitochondrial dysfunction, linear lactate pyruvate, inhibitor, ANCOVA, rotenone, shift · block
oscr:b:X1HPPGJYHN9KV-K· Report a mismatch - [12] § Results › Linear parameters and the hyperbolic consequence › Characterisation ↔ 5_linear_LP_models.R, lines 987 to 1026 · score 0.54 · median R2, negative intercept, positive intercept, Stratification, IQR, linear · block
oscr:b:CZ4EMYEXF9CZM-R· Report a mismatch - [13] § Results › Explaining metabolic polarisation ↔ 7_never_always_LPR25.R, lines 1 to 78 · score 0.54 · discarded segments, linear parameters, threshold, retained segments, filtering, datapoints · block
oscr:b:JMERCWB0991RX-S· Report a mismatch - [14] § Results › Outcome association ↔ 9_outcome.R, lines 1 to 81 · score 0.53 · worse outcome, linear contrast, steeper, Death, CI, patients · block
oscr:b:E2PKTJR09HRHV-I· Report a mismatch - [15] § Results › Substrate delivery and LPR fidelity ↔ 8_substrate_lpr_fidelity.R, lines 61 to 118 · score 0.52 · anaerobic activity, LPR glucose, substrate, fidelity, tracks, validating · block
oscr:b:YAEMJKBV0F950-A· Report a mismatch
Paper
Loaded from Europe PMC by your browser; OSCR does not store it: doi.org · Europe PMC
The paper is loaded when this pane is shown.
The authors' code
R · 572 lines · 25 KB · MIT · 4 matches
- # ==============================================================================
- # Script: 8_substrate_lpr_fidelity.R
- # Manuscript relevance: 3.5, Fig. 5
- # ==============================================================================
- # PURPOSE:
- # Elucidate the relationship between substrate delivery (glucose), the linear
- # gradient (m), and the LPR at within-segment and between-segment levels.
- #
- # INPUT:
- # - 1_output/2_linear_segmentation/1_results.csv: Segment-annotated data
- #
- # OUTPUT:
- # - 1_output/8_substrate_lpr_fidelity/__execution_time.csv: Runtime log
- # - 1_output/8_substrate_lpr_fidelity/_1_model_summaries.txt: Full lmer summaries
- # - 1_output/8_substrate_lpr_fidelity/_1_stats.csv: Model coefficients and statistics
- # - 1_output/8_substrate_lpr_fidelity/0_notes.txt: Figure legend
- # - 1_output/8_substrate_lpr_fidelity/1_figure.png: 4-panel figure
- #
- # DATA FILTERS:
- # - Retained segments only: p6e_m > 0 (positive gradient)
- # - Observations with non-NA glucose
- #
- # ANALYSES:
- # ------------------------------------------------------------------------------
- # Seven LME models. All use Satterthwaite approximation for t-tests.
- # Models 1, 5 produce no figure panels.
- #
- # Model 1: Pyruvate ~ Glucose (POINT-LEVEL, no panel)
- # - Establishes pyruvate–glucose association; reported in text.
- # Unit: Point-level (individual datapoints)
- # Model: pyruvate ~ glucose + (1|patient/p6e_seg_index)
- # Weighting: None
- # Hypothesis: β_glucose > 0 (one-sided)
- # Test: Satterthwaite approximation
- #
- # Model 2: LPR ~ 1/Glucose in Type Pb segments (POINT-LEVEL → Panel A)
- # - Within-segment: LPR vs glucose when b > 0.
- # Modelled as lpr ~ inv_glucose to capture hyperbolic shape;
- # plotted on glucose x-axis.
- # Unit: Point-level, Type Pb segments only
- # Model: lpr ~ inv_glucose + (1|patient/p6e_seg_index)
- # Weighting: None
- # Hypothesis: β > 0 (one-sided; theory: b > 0 → LPR ↓ as glucose ↑)
- # Test: Satterthwaite approximation
- #
- # Model 3: LPR ~ 1/Glucose in Type Nb segments (POINT-LEVEL → Panel B)
- # - Within-segment: LPR vs glucose when b < 0.
- # Unit: Point-level, Type Nb segments only
- # Model: lpr ~ inv_glucose + (1|patient/p6e_seg_index)
- # Weighting: None
- # Hypothesis: β < 0 (one-sided; theory: b < 0 → LPR ↑ as glucose ↑)
- # Test: Satterthwaite approximation
- #
- # Model 4: LPR ~ m (SEGMENT-LEVEL → Panel C)
- # - How faithfully does LPR track the underlying gradient?
- # Unit: Segment-level (median LPR, gradient per segment)
- # Model: median_lpr ~ m + (1|patient)
- # Weighting: n_points (segment size)
- # Hypothesis: β > 0 (one-sided); theoretical β = 1 if LPR = m
- # Test: Satterthwaite approximation
- #
- # Model 5: b ~ m (SEGMENT-LEVEL, no panel)
- # - Exploratory: quantifies the m–b coupling
- # Unit: Segment-level
- # Model: b ~ m + (1|patient)
- # Weighting: n_points (segment size)
- # Hypothesis: Two-sided (exploratory)
- # Test: Satterthwaite approximation
- #
- # Model 6: m ~ Glucose (SEGMENT-LEVEL → Panel D)
- # - Is substrate availability associated with anaerobic activity?
- # Unit: Segment-level (gradient, median glucose per segment)
- # Model: m ~ glucose + (1|patient)
- # Weighting: n_points (segment size)
- # Hypothesis: β < 0 (higher glucose → lower m; one-sided)
- # Test: Satterthwaite approximation
- #
- # Model 7: LPR ~ Glucose (SEGMENT-LEVEL → Panel D)
- # - Between-segment: does LPR track glucose at the segment level?
- # Unit: Segment-level (median LPR, median glucose per segment)
- # Model: median_lpr ~ glucose + (1|patient)
- # Weighting: n_points (segment size)
- # Hypothesis: Two-sided (exploratory)
- # Test: Satterthwaite approximation
- #
- # Figure panels:
- # Panel A: Within-segment LPR vs glucose, Type Pb (Model 2; hyperbolic fit)
- # Panel B: Within-segment LPR vs glucose, Type Nb (Model 3; hyperbolic fit)
- # Panel C: Between-segment median LPR vs m (Model 4)
- # Panel D: Between-segment m (Model 6) and LPR (Model 7) vs median glucose
- #
- # MULTIPLE COMPARISONS CORRECTION:
- # Applied only to Models 2 and 3 (Benjamini-Hochberg FDR, n=2). These two models jointly test
- # the overarching hyperbolic hypothesis (LPR vs 1/Glucose) across two mutually
- # exclusive subsets (Pb and Nb). The remaining models are uncorrected because they
- # address distinct, independent physiological questions:
- # - Model 1: Validates the fundamental Pyruvate-Glucose link.
- # - Model 4: Independent assessment of LPR-m association.
- # - Model 5: Exploratory quantification of the m-b coupling.
- # - Model 6: Tests the primary hypothesis linking m to substrate availability.
- # - Model 7: Exploratory comparison of LPR tracking vs m tracking.
- # ==============================================================================
- suppressPackageStartupMessages({
- library(here)
- library(dplyr)
- library(readr)
- library(ggplot2)
- library(lme4)
- library(lmerTest)
- library(patchwork)
- })
- source(here::here("_shared", "notes.R"))
- cat("\n\nStarting 8_substrate_lpr_fidelity.R\n")
- start_time <- Sys.time()
- save_plot_portable <- function(filename, plot, ...) {
- if (isTRUE(capabilities("cairo"))) {
- ggsave(filename = filename, plot = plot, type = "cairo", ...)
- } else {
- warning(sprintf("Cairo backend unavailable; saving %s with the default device.", basename(filename)))
- ggsave(filename = filename, plot = plot, ...)
- }
- }
- safe_divide <- function(numerator, denominator, tol = 1e-12, fill = NA_real_) {
- result <- numerator / denominator
- invalid <- !is.finite(result) | !is.finite(denominator) | abs(denominator) <= tol
- result[invalid] <- fill
- result
- }
- get_lmer_pred_se <- function(model, newdata) {
- fit <- predict(model, newdata = newdata, re.form = NA, allow.new.levels = TRUE)
- X <- model.matrix(delete.response(terms(model)), newdata)
- vcov_mat <- as.matrix(vcov(model))
- se <- sqrt(rowSums((X %*% vcov_mat) * X))
- list(fit = fit, se = se)
- }
- extract_coef <- function(model, predictor, one_sided = FALSE, expected_sign = 1) {
- coefs <- summary(model)$coefficients
- beta <- coefs[predictor, "Estimate"]
- se <- coefs[predictor, "Std. Error"]
- t_val <- coefs[predictor, "t value"]
- p_two <- coefs[predictor, "Pr(>|t|)"]
- if (one_sided) {
- p_val <- if ((expected_sign > 0 && beta > 0) || (expected_sign < 0 && beta < 0)) p_two / 2 else 1 - p_two / 2
- } else {
- p_val <- p_two
- }
- list(beta = beta, se = se, t_value = t_val, p_value = p_val)
- }
- format_p_text <- function(p) if (p < 0.001) "p < 0.001" else sprintf("p = %.3f", p)
- # ==============================================================================
- # Input / Output
- # ==============================================================================
- INPUT_FILE <- here::here("1_output", "2_linear_segmentation", "1_results.csv")
- if (!file.exists(INPUT_FILE)) {
- cat("\n\nError: Results file not found. Run 2_linear_segmentation.py first.\n")
- stop()
- }
- OUTPUT_DIR <- here::here("1_output", "8_substrate_lpr_fidelity")
- if (!dir.exists(OUTPUT_DIR)) dir.create(OUTPUT_DIR, recursive = TRUE)
- # ==============================================================================
- # Data Loading and Preparation
- # ==============================================================================
- raw_df <- read_csv(INPUT_FILE, show_col_types = FALSE, progress = FALSE)
- retained_df <- raw_df %>%
- filter(is.finite(p6e_m), p6e_m > 0)
- # Point-level data
- point_df <- retained_df %>%
- filter(!is.na(glucose)) %>%
- select(patient, p6e_seg_index, m = p6e_m, b = p6e_b, glucose, pyruvate, lpr) %>%
- mutate(
- type_b = case_when(b > 0 ~ "Pb", b < 0 ~ "Nb", TRUE ~ "Zb"),
- inv_glucose = safe_divide(1, glucose)
- )
- point_pb <- point_df %>% filter(type_b == "Pb")
- point_nb <- point_df %>% filter(type_b == "Nb")
- # Segment-level data
- segment_df <- point_df %>%
- group_by(patient, p6e_seg_index) %>%
- summarise(
- median_lpr = median(lpr, na.rm = TRUE),
- m = first(m),
- b = first(b),
- glucose = median(glucose, na.rm = TRUE),
- n_points = n(),
- .groups = "drop"
- )
- n_points_total <- nrow(point_df)
- n_segments <- nrow(segment_df)
- n_patients <- n_distinct(segment_df$patient)
- n_points_pb <- nrow(point_pb)
- n_seg_pb <- n_distinct(paste(point_pb$patient, point_pb$p6e_seg_index))
- n_pat_pb <- n_distinct(point_pb$patient)
- n_points_nb <- nrow(point_nb)
- n_seg_nb <- n_distinct(paste(point_nb$patient, point_nb$p6e_seg_index))
- n_pat_nb <- n_distinct(point_nb$patient)
- # ==============================================================================
- # Statistical Models
- # ==============================================================================
- # --- Model 1: Pyruvate (mM) ~ Glucose (POINT-LEVEL, no panel) ---
- model_1 <- lmer(pyruvate ~ glucose + (1|patient/p6e_seg_index), data = point_df)
- stats_1 <- extract_coef(model_1, "glucose", one_sided = TRUE, expected_sign = 1)
- resid_sd_pyr <- sigma(model_1)
- # --- Model 2: LPR ~ 1/Glucose in Type Pb (POINT-LEVEL → Panel A) ---
- model_2 <- lmer(lpr ~ inv_glucose + (1|patient/p6e_seg_index), data = point_pb)
- stats_2 <- extract_coef(model_2, "inv_glucose", one_sided = TRUE, expected_sign = 1)
- # --- Model 3: LPR ~ 1/Glucose in Type Nb (POINT-LEVEL → Panel B) ---
- model_3 <- lmer(lpr ~ inv_glucose + (1|patient/p6e_seg_index), data = point_nb)
- stats_3 <- extract_coef(model_3, "inv_glucose", one_sided = TRUE, expected_sign = -1)
- # Apply Benjamini-Hochberg FDR correction for Models 2 & 3 (family of 2 tests for hyperbolic hypothesis)
- p_adj_23 <- p.adjust(c(stats_2$p_value, stats_3$p_value), method = "BH")
- stats_2$p_value <- p_adj_23[1]
- stats_3$p_value <- p_adj_23[2]
- # --- Model 4: LPR ~ m (SEGMENT-LEVEL → Panel C) ---
- model_4 <- lmer(median_lpr ~ m + (1|patient), data = segment_df, weights = n_points,
- control = lmerControl(check.conv.singular = .makeCC(action = "ignore", tol = 1e-4)))
- stats_4 <- extract_coef(model_4, "m", one_sided = TRUE, expected_sign = 1)
- # --- Model 5: b ~ m (SEGMENT-LEVEL, no panel) ---
- model_5 <- lmer(b ~ m + (1|patient), data = segment_df, weights = n_points,
- control = lmerControl(check.conv.singular = .makeCC(action = "ignore", tol = 1e-4)))
- stats_5 <- extract_coef(model_5, "m", one_sided = FALSE)
- # --- Model 6: m ~ Glucose (SEGMENT-LEVEL → Panel D) ---
- model_6 <- lmer(m ~ glucose + (1|patient), data = segment_df, weights = n_points,
- control = lmerControl(check.conv.singular = .makeCC(action = "ignore", tol = 1e-4)))
- stats_6 <- extract_coef(model_6, "glucose", one_sided = TRUE, expected_sign = -1)
- # --- Model 7: LPR ~ Glucose (SEGMENT-LEVEL → Panel D) ---
- model_7 <- lmer(median_lpr ~ glucose + (1|patient), data = segment_df, weights = n_points,
- control = lmerControl(check.conv.singular = .makeCC(action = "ignore", tol = 1e-4)))
- stats_7 <- extract_coef(model_7, "glucose", one_sided = FALSE)
- p_text_1 <- format_p_text(stats_1$p_value)
- p_text_2 <- format_p_text(stats_2$p_value)
- p_text_3 <- format_p_text(stats_3$p_value)
- p_text_4 <- format_p_text(stats_4$p_value)
- p_text_5 <- format_p_text(stats_5$p_value)
- p_text_6 <- format_p_text(stats_6$p_value)
- p_text_7 <- format_p_text(stats_7$p_value)
- # Save model summaries
- sink(file.path(OUTPUT_DIR, "_1_model_summaries.txt"))
- cat("=======================================================================\n")
- cat("Model 1: Pyruvate (mM) ~ Glucose [No panel] (POINT-LEVEL)\n")
- cat("Establishes pyruvate–glucose association; reported in text.\n")
- cat(sprintf("Residual SD = %.5f mM (%.1f μM).\n", resid_sd_pyr, resid_sd_pyr * 1000))
- cat(sprintf("N points = %d, N segments = %d, N patients = %d\n",
- n_points_total, n_segments, n_distinct(point_df$patient)))
- cat("=======================================================================\n")
- print(summary(model_1))
- cat("\n\n")
- cat("=======================================================================\n")
- cat("Model 2: LPR ~ 1/Glucose in Type Pb [Panel A] (POINT-LEVEL)\n")
- cat("Within-segment: LPR–glucose association when b > 0.\n")
- cat("Modelled as lpr ~ inv_glucose; plotted on glucose axis (hyperbolic shape).\n")
- cat(sprintf("N points = %d, N segments = %d, N patients = %d\n",
- n_points_pb, n_seg_pb, n_pat_pb))
- cat(sprintf("Benjamini-Hochberg FDR corrected p-value (n=2): %g\n", stats_2$p_value))
- cat("=======================================================================\n")
- print(summary(model_2))
- cat("\n\n")
- cat("=======================================================================\n")
- cat("Model 3: LPR ~ 1/Glucose in Type Nb [Panel B] (POINT-LEVEL)\n")
- cat("Within-segment: LPR–glucose association when b < 0.\n")
- cat("Modelled as lpr ~ inv_glucose; plotted on glucose axis (hyperbolic shape).\n")
- cat(sprintf("N points = %d, N segments = %d, N patients = %d\n",
- n_points_nb, n_seg_nb, n_pat_nb))
- cat(sprintf("Benjamini-Hochberg FDR corrected p-value (n=2): %g\n", stats_3$p_value))
- cat("=======================================================================\n")
- print(summary(model_3))
- cat("\n\n")
- cat("=======================================================================\n")
- cat("Model 4: Median LPR ~ m [Panel C] (SEGMENT-LEVEL)\n")
- cat("How faithfully does LPR track the underlying gradient?\n")
- cat("Theoretical β = 1 if LPR perfectly tracked m.\n")
- cat(sprintf("N segments = %d, N patients = %d\n", n_segments, n_patients))
- cat("=======================================================================\n")
- print(summary(model_4))
- cat("\n\n")
- cat("=======================================================================\n")
- cat("Model 5: b ~ m [No panel] (SEGMENT-LEVEL)\n")
- cat("Exploratory: quantifies the m–b coupling.\n")
- cat(sprintf("N segments = %d, N patients = %d\n", n_segments, n_patients))
- cat("=======================================================================\n")
- print(summary(model_5))
- cat("\n\n")
- cat("=======================================================================\n")
- cat("Model 6: m ~ Glucose [Panel D] (SEGMENT-LEVEL)\n")
- cat("Between-segment: is substrate availability associated with gradient?\n")
- cat(sprintf("N segments = %d, N patients = %d\n", n_segments, n_patients))
- cat("=======================================================================\n")
- print(summary(model_6))
- cat("\n\n")
- cat("=======================================================================\n")
- cat("Model 7: Median LPR ~ Glucose [Panel D] (SEGMENT-LEVEL)\n")
- cat("Between-segment: does segment-level LPR track glucose?\n")
- cat(sprintf("N segments = %d, N patients = %d\n", n_segments, n_patients))
- cat("=======================================================================\n")
- print(summary(model_7))
- sink()
- # ==============================================================================
- # Visualization (4-Panel Figure)
- # ==============================================================================
- common_theme <- theme_minimal() +
- theme(
- text = element_text(size = 8, colour = "black"),
- axis.text = element_text(size = 6, colour = "black"),
- axis.title = element_text(size = 7, colour = "black"),
- plot.tag = element_text(size = 9, face = "bold", colour = "black", hjust = 1, vjust = -1),
- plot.tag.position = c(0.02, 0.99),
- panel.border = element_rect(colour = "black", fill = NA, linewidth = 0.5),
- panel.grid.minor = element_blank(),
- plot.background = element_rect(fill = "white", color = NA),
- panel.background = element_rect(fill = "white", color = NA),
- legend.position = "none",
- plot.margin = margin(t = 3, r = 3, b = 0, l = 3, unit = "mm")
- )
- # Viewports (point-level for A–B, segment-level for C–D)
- glucose_range_pb <- quantile(point_pb$glucose, c(0.025, 0.975), na.rm = TRUE)
- glucose_range_nb <- quantile(point_nb$glucose, c(0.025, 0.975), na.rm = TRUE)
- glucose_range_seg <- quantile(segment_df$glucose, c(0.025, 0.975), na.rm = TRUE)
- m_range <- quantile(segment_df$m, c(0.025, 0.975), na.rm = TRUE)
- glucose_seq_pb <- seq(glucose_range_pb[1], glucose_range_pb[2], length.out = 100)
- glucose_seq_nb <- seq(glucose_range_nb[1], glucose_range_nb[2], length.out = 100)
- glucose_seq_seg <- seq(glucose_range_seg[1], glucose_range_seg[2], length.out = 100)
- m_seq <- seq(m_range[1], m_range[2], length.out = 100)
- # --- Panel A: LPR ~ 1/Glucose within Pb (Model 2, plotted on glucose axis) ---
- pred_a <- data.frame(inv_glucose = safe_divide(1, glucose_seq_pb))
- res_a <- get_lmer_pred_se(model_2, pred_a)
- pred_a$glucose <- glucose_seq_pb
- pred_a$y <- res_a$fit
- pred_a$ymin <- pred_a$y - 1.96 * res_a$se
- pred_a$ymax <- pred_a$y + 1.96 * res_a$se
- ylim_a <- range(c(pred_a$ymin[1], pred_a$ymax[1],
- pred_a$ymin[nrow(pred_a)], pred_a$ymax[nrow(pred_a)]))
- p_a <- ggplot() +
- geom_ribbon(data = pred_a, aes(x = glucose, ymin = ymin, ymax = ymax),
- fill = "#44AA99", alpha = 0.2) +
- geom_line(data = pred_a, aes(x = glucose, y = y),
- color = "#44AA99", linewidth = 1) +
- coord_cartesian(xlim = glucose_range_pb, ylim = ylim_a) +
- scale_x_continuous(breaks = scales::pretty_breaks(n = 5)) +
- scale_y_continuous(breaks = scales::pretty_breaks(n = 5)) +
- labs(tag = "A", x = "Glucose (mM)", y = "LPR") +
- common_theme
- # --- Panel B: LPR ~ 1/Glucose within Nb (Model 3, plotted on glucose axis) ---
- pred_b <- data.frame(inv_glucose = safe_divide(1, glucose_seq_nb))
- res_b <- get_lmer_pred_se(model_3, pred_b)
- pred_b$glucose <- glucose_seq_nb
- pred_b$y <- res_b$fit
- pred_b$ymin <- pred_b$y - 1.96 * res_b$se
- pred_b$ymax <- pred_b$y + 1.96 * res_b$se
- ylim_b <- range(c(pred_b$ymin[1], pred_b$ymax[1],
- pred_b$ymin[nrow(pred_b)], pred_b$ymax[nrow(pred_b)]))
- p_b <- ggplot() +
- geom_ribbon(data = pred_b, aes(x = glucose, ymin = ymin, ymax = ymax),
- fill = "#DDCC77", alpha = 0.2) +
- geom_line(data = pred_b, aes(x = glucose, y = y),
- color = "#DDCC77", linewidth = 1) +
- coord_cartesian(xlim = glucose_range_nb, ylim = ylim_b) +
- scale_x_continuous(breaks = scales::pretty_breaks(n = 5)) +
- scale_y_continuous(breaks = scales::pretty_breaks(n = 5)) +
- labs(tag = "B", x = "Glucose (mM)", y = "LPR") +
- common_theme
- # --- Panel C: LPR ~ m (Model 4, segment-level) ---
- pred_c <- data.frame(m = m_seq)
- res_c <- get_lmer_pred_se(model_4, pred_c)
- pred_c$y <- res_c$fit
- pred_c$ymin <- pred_c$y - 1.96 * res_c$se
- pred_c$ymax <- pred_c$y + 1.96 * res_c$se
- ref_c <- data.frame(m = m_range, lpr = m_range)
- ylim_c <- range(c(pred_c$ymin[1], pred_c$ymax[1],
- pred_c$ymin[nrow(pred_c)], pred_c$ymax[nrow(pred_c)],
- m_range))
- p_c <- ggplot() +
- geom_line(data = ref_c, aes(x = m, y = lpr),
- linetype = "dashed", color = "black", linewidth = 0.6) +
- geom_ribbon(data = pred_c, aes(x = m, ymin = ymin, ymax = ymax),
- fill = "#CC3311", alpha = 0.2) +
- geom_line(data = pred_c, aes(x = m, y = y),
- color = "#CC3311", linewidth = 1) +
- coord_cartesian(xlim = m_range, ylim = ylim_c) +
- scale_x_continuous(breaks = scales::pretty_breaks(n = 5)) +
- scale_y_continuous(breaks = scales::pretty_breaks(n = 5)) +
- labs(tag = "C", x = expression(italic(m)), y = "LPR") +
- common_theme
- # --- Panel D: m AND LPR ~ Glucose (Models 6 & 7, segment-level) ---
- pred_d_m <- data.frame(glucose = glucose_seq_seg)
- res_d_m <- get_lmer_pred_se(model_6, pred_d_m)
- pred_d_m$y <- res_d_m$fit
- pred_d_m$ymin <- pred_d_m$y - 1.96 * res_d_m$se
- pred_d_m$ymax <- pred_d_m$y + 1.96 * res_d_m$se
- pred_d_lpr <- data.frame(glucose = glucose_seq_seg)
- res_d_lpr <- get_lmer_pred_se(model_7, pred_d_lpr)
- pred_d_lpr$y <- res_d_lpr$fit
- pred_d_lpr$ymin <- pred_d_lpr$y - 1.96 * res_d_lpr$se
- pred_d_lpr$ymax <- pred_d_lpr$y + 1.96 * res_d_lpr$se
- ylim_d <- range(c(pred_d_m$ymin[1], pred_d_m$ymax[1],
- pred_d_m$ymin[nrow(pred_d_m)], pred_d_m$ymax[nrow(pred_d_m)],
- pred_d_lpr$ymin[1], pred_d_lpr$ymax[1],
- pred_d_lpr$ymin[nrow(pred_d_lpr)], pred_d_lpr$ymax[nrow(pred_d_lpr)]))
- p_d <- ggplot() +
- geom_ribbon(data = pred_d_lpr, aes(x = glucose, ymin = ymin, ymax = ymax),
- fill = "#E69F00", alpha = 0.2) +
- geom_line(data = pred_d_lpr, aes(x = glucose, y = y),
- color = "#E69F00", linewidth = 1) +
- geom_ribbon(data = pred_d_m, aes(x = glucose, ymin = ymin, ymax = ymax),
- fill = "#0072B2", alpha = 0.2) +
- geom_line(data = pred_d_m, aes(x = glucose, y = y),
- color = "#0072B2", linewidth = 1) +
- coord_cartesian(xlim = glucose_range_seg, ylim = ylim_d) +
- scale_x_continuous(breaks = scales::pretty_breaks(n = 5)) +
- scale_y_continuous(breaks = scales::pretty_breaks(n = 5)) +
- labs(tag = "D", x = "Glucose (mM)", y = "Value") +
- common_theme
- # Assemble 4-panel figure
- fig <- p_a + p_b + p_c + p_d +
- plot_layout(nrow = 1)
- save_plot_portable(
- filename = file.path(OUTPUT_DIR, "1_figure.png"),
- plot = fig,
- width = 185, height = 55, units = "mm", dpi = 300
- )
- # ==============================================================================
- # Summary Statistics CSV
- # ==============================================================================
- summary_stats <- data.frame(
- Metric = c(
- "N_points", "N_segments", "N_patients",
- "N_points_Pb", "N_segments_Pb", "N_patients_Pb",
- "N_points_Nb", "N_segments_Nb", "N_patients_Nb",
- "Model_1_beta_glucose_mM_per_mM", "Model_1_beta_glucose_uM_per_mM",
- "Model_1_SE_uM_per_mM", "Model_1_p_value", "Model_1_resid_SD_mM",
- "Model_2_beta_inv_glucose_Pb", "Model_2_SE_Pb", "Model_2_p_value_bh_Pb",
- "Model_3_beta_inv_glucose_Nb", "Model_3_SE_Nb", "Model_3_p_value_bh_Nb",
- "Model_4_beta_m", "Model_4_SE", "Model_4_p_value",
- "Model_5_beta_m_bm", "Model_5_SE_bm", "Model_5_p_value_bm",
- "Model_6_beta_glucose", "Model_6_SE", "Model_6_p_value",
- "Model_7_beta_glucose_lpr", "Model_7_SE_lpr", "Model_7_p_value_lpr"
- ),
- Value = c(
- n_points_total, n_segments, n_patients,
- n_points_pb, n_seg_pb, n_pat_pb,
- n_points_nb, n_seg_nb, n_pat_nb,
- stats_1$beta, stats_1$beta * 1000,
- stats_1$se * 1000, stats_1$p_value, resid_sd_pyr,
- stats_2$beta, stats_2$se, stats_2$p_value,
- stats_3$beta, stats_3$se, stats_3$p_value,
- stats_4$beta, stats_4$se, stats_4$p_value,
- stats_5$beta, stats_5$se, stats_5$p_value,
- stats_6$beta, stats_6$se, stats_6$p_value,
- stats_7$beta, stats_7$se, stats_7$p_value
- )
- )
- write_csv(summary_stats, file.path(OUTPUT_DIR, "_1_stats.csv"))
- # ==============================================================================
- # Completion Log
- # ==============================================================================
- end_time <- Sys.time()
- execution_time <- as.numeric(difftime(end_time, start_time, units = "secs"))
- write_csv(
- data.frame(execution_time_seconds = execution_time),
- file.path(OUTPUT_DIR, "__execution_time.csv")
- )
- cat("\n\n8_substrate_lpr_fidelity.R complete.\n\n")
- # ==============================================================================
- # Figure Legend
- # ==============================================================================
- legends <- list(
- list(
- target = "Figure 1 (1_figure.png)",
- caption = sprintf(
- paste0(
- "Glucose associations with LPR and gradient (m). ",
- "(A) Within-segment LPR vs glucose for Type Pb segments (b > 0; %d segments, %d patients; ",
- "β_1/Glucose = %.2f mM, %s). ",
- "(B) Within-segment LPR vs glucose for Type Nb segments (b < 0; %d segments, %d patients; ",
- "β_1/Glucose = %.2f mM, %s). ",
- "Within-segment models use 1/Glucose as predictor to capture the hyperbolic LPR–glucose ",
- "relationship (LPR = m + b/P); predictions are plotted on the glucose x-axis. ",
- "P-values for within-segment models (A, B) are Benjamini-Hochberg FDR corrected (n=2). ",
- "(C) Segment-level median LPR vs gradient m (β = %.3f, %s); dashed black line indicates the ",
- "theoretical β = 1 relationship expected if LPR perfectly tracked the gradient. ",
- "(D) Segment-level gradient (m, blue; β = %.3f mM⁻¹, %s) and median LPR (orange; β = %.3f mM⁻¹, %s) ",
- "vs median glucose. ",
- "Lines = LME fixed effects; shaded = 95%% CI. ",
- "A–B: nested random intercepts (segment within patient). ",
- "C–D: patient random intercepts, weighted by segment size. ",
- "Sample: %d segments, %d patients."
- ),
- n_seg_pb, n_pat_pb, stats_2$beta, p_text_2,
- n_seg_nb, n_pat_nb, stats_3$beta, p_text_3,
- stats_4$beta, p_text_4,
- stats_6$beta, p_text_6,
- stats_7$beta, p_text_7,
- n_segments, n_patients
- ),
- abbreviations = "b, linear intercept; BH, Benjamini-Hochberg; CI, confidence interval; FDR, false discovery rate; LME, linear mixed-effects model; LPR, lactate/pyruvate ratio; m, linear gradient; Nb, negative intercept (b < 0); Pb, positive intercept (b > 0)"
- ),
- list(
- target = "Text-only results (no panel)",
- caption = sprintf(
- paste0(
- "Model 1 (pyruvate ~ glucose, point-level): β = %.2f μM/mM (%s); ",
- "residual SD = %.1f μM. %d datapoints, %d segments, %d patients. ",
- "Model 5 (b ~ m, segment-level): β = %.4f mM (%s). ",
- "%d segments, %d patients."
- ),
- stats_1$beta * 1000, p_text_1,
- resid_sd_pyr * 1000, n_points_total, n_segments, n_patients,
- stats_5$beta, p_text_5,
- n_segments, n_patients
- ),
- abbreviations = "b, linear intercept; LPR, lactate/pyruvate ratio; m, linear gradient; SD, standard deviation"
- )
- )
- writeLines(build_notes(legends, title = "8_substrate_lpr_fidelity"), file.path(OUTPUT_DIR, "0_notes.txt"))
8_substrate_lpr_fidelity.R at commit 7cc6e72, under MIT · at the source
Overview
- Department of Clinical Neurosciences, University of Cambridge, Cambridge, UK
- Section of Neurosurgery, Department of Medical Sciences, Uppsala University Hospital, Uppsala, Sweden
- Wellcome-MRC Cambridge Stem Cell Institute, University of Cambridge, Cambridge, UK
- Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden
Abstract
Cerebral microdialysis (CMD) after traumatic brain injury (TBI) has focused on lactate/
Reproduced under the paper's license (CC BY), from the paper cited above.
Repository
Its files are read in the Code ↔ Paper reader above, with 15 matches between paragraphs and lines of code.
msb-jr/LPG
7cc6e72aec21e4b971deb857b3e2f671a79b399e, 9 September 2026Availability: 1 check, the latest on 27 September 2026: the link answers
- 27 September 2026: the link answers
18 files
- 10_in_vitro.R, R, 263 lines, 3 matches
- 1_pre-processing.py, Python, 163 lines
- 2_linear_segmentation.py
, Python, 636 lines - 3_segmentation_visual.py
, Python, 1,094 lines - 4_segmentation_stats.R, R, 246 lines, 1 match
- 5_linear_LP_models.R, R, 1,131 lines, 2 matches
- 6_hyperbolic_LPR_error.R
, R, 1,214 lines, 1 match - 7_never_always_LPR25.R, R, 577 lines, 2 matches
- 8_substrate_lpr_fidelity
.R , R, 572 lines, 4 matches - 9_outcome.R, R, 410 lines, 2 matches
- _shared/
__init__.py , Python, 5 lines - _shared/
notes.R , R, 111 lines - _shared/
notes.py , Python, 117 lines - extra_linear_to_hyperbol
ic.py , Python, 267 lines - extra_lpr_m_schematic.py
, Python, 341 lines - renv/
activate.R , R, 1,334 lines - LICENSE.txt, License, 9 lines
- README.md, Text, 153 lines
Statistical analysis and code availability
Multiple statistical approaches were used, including OLS regression, Fisher’s z-transformed correlation comparisons, model comparison via Akaike Information Criterion (AIC), linear mixed-effects (LME) models with nested random effects, nonparametric tests for group comparisons with Benjamini-Hochberg correction for multiple testing, linear contrasts via estimated marginal means (EMMs), and analysis of covariance (ANCOVA) for comparing in vitro regression lines. EMMs represent the model-estimated typical values within each group, adjusted for the model’s random effects and weighting structure. All analyses were performed using R (v4.5.2; https://
Reproduced under the paper's license (CC BY), from the paper cited above.
Tracing map
Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.
What the map holds:
- 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
- 16 scripts, each with its path and the digest of its content;
- 15 matches between paragraphs of the paper and lines of the code (method lexical-v1);
- neither the text of the paper nor the code itself.
Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.
Data
No dataset and no data link were found in the paper.
Versions
The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.
Version 1, 27 September 2026: the first record
Recorded: type, language, journal, pages, dates, 14 authors, 5 keywords, 2 funders, 24 references.
Cite
This paper
Baker, M. S., Heihre, J. M., Kuliński, T., Smith, C. A., Barlas, B., Zimphango, C., Timofeev, I., Baranes, K., Carpenter, K. L. H., Kotter, M., Guilfoyle, M. R., Rostami, E., Hutchinson, P. J., & Helmy, A. (2026). The microdialysis-derived lactate-pyruvate gradient indicates anaerobic activity after brain injury. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism, 0271678X261485404. https://
BibTeX
@article{baker2026microd
author = {Baker, Michael S and Heihre, Joshua M and Kuliński, Tomasz and Smith, Claudia A and Barlas, Berfin and Zimphango, Chisomo and Timofeev, Ivan and Baranes, Koby and Carpenter, Keri L H and Kotter, Mark and Guilfoyle, Mathew R and Rostami, Elham and Hutchinson, Peter J and Helmy, Adel},
title = {{The microdialysis-derived lactate-pyruvate gradient indicates anaerobic activity after brain injury}},
journal = {Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism},
year = {2026},
month = aug,
pages = {0271678X261485404},
publisher = {SAGE Publications},
issn = {0271-678X},
doi = {10.1177/
url = {https://
pmid = {42644382},
pmcid = {PMC13558473}
}
RIS
TY - JOUR
AU - Baker, Michael S
AU - Heihre, Joshua M
AU - Kuliński, Tomasz
AU - Smith, Claudia A
AU - Barlas, Berfin
AU - Zimphango, Chisomo
AU - Timofeev, Ivan
AU - Baranes, Koby
AU - Carpenter, Keri L H
AU - Kotter, Mark
AU - Guilfoyle, Mathew R
AU - Rostami, Elham
AU - Hutchinson, Peter J
AU - Helmy, Adel
TI - The microdialysis-derived lactate-pyruvate gradient indicates anaerobic activity after brain injury
T2 - Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism
J2 - J Cereb Blood Flow Metab
PY - 2026
DA - 2026/
SP - 0271678X261485404
SN - 0271-678X
PB - SAGE Publications
DO - 10.1177/
UR - https://
LA - en
ER -
CSL-JSON
{
"id": "10.1177/
"type": "article-journal",
"title": "The microdialysis-derived lactate-pyruvate gradient indicates anaerobic activity after brain injury",
"container-title": "Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism",
"author": [
{
"family": "Baker",
"given": "Michael S"
},
{
"family": "Heihre",
"given": "Joshua M"
},
{
"family": "Kuliński",
"given": "Tomasz"
},
{
"family": "Smith",
"given": "Claudia A"
},
{
"family": "Barlas",
"given": "Berfin"
},
{
"family": "Zimphango",
"given": "Chisomo"
},
{
"family": "Timofeev",
"given": "Ivan"
},
{
"family": "Baranes",
"given": "Koby"
},
{
"family": "Carpenter",
"given": "Keri L H"
},
{
"family": "Kotter",
"given": "Mark"
},
{
"family": "Guilfoyle",
"given": "Mathew R"
},
{
"family": "Rostami",
"given": "Elham"
},
{
"family": "Hutchinson",
"given": "Peter J"
},
{
"family": "Helmy",
"given": "Adel"
}
],
"container-title-short":
"page": "0271678X261485404",
"DOI": "10.1177/
"PMID": "42644382",
"PMCID": "PMC13558473",
"ISSN": "0271-678X",
"publisher": "SAGE Publications",
"URL": "https://
"language": "en",
"issued": {
"date-parts": [
[
2026,
8,
26
]
]
}
}
The tracing map gets a citation of its own once an author has validated it and it has a DOI.
Similar papers
The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.
- [1] doi:10.1002/hbm.70512 [code]
- Precision Imaging for Intraindividual Investigation of the Reward Response.Journal: Human brain mappingIn common: lmerTest, lme4, patchwork, 5 other tools, 1 reference
- [2] doi:10.1093/braincomms/fcaf096 [code]
- Influence of apolipoprotein E genotype on the proteomic profile in cerebral microdialysis after human severe traumatic brain injury: a prospective observational study.Journal: Brain communicationsIn common: tidyverse, 2 references, author Elham Rostami
- [3] doi:10.1186/s13024-024-00773-1 [code]
- Human VCP mutant ALS/
FTD microglia display immune and lysosomal phenotypes independently of GPNMB. Journal: Molecular neurodegenerationIn common: patchwork, tidyverse, cellular / molecular, author Adel Helmy - [4] doi:10.1093/braincomms/fcag239 [code]
- Extracellular matrix remodelling in degenerative cervical myelopathy.Journal: Brain communicationsIn common: ggplot2, tidyverse, author Koby Baranes
- [5] doi:10.3390/cells11020241 [code]
- Expression of Lineage Transcription Factors Identifies Differences in Transition States of Induced Human Oligodendrocyte Differentiation.Journal: CellsIn common: patchwork, ggplot2, tidyverse, 3 other tools, cellular / molecular, 1 reference
- [6] doi:10.1038/s42003-025-08194-6 [code]
- Drug treatment alters performance in a neural microphysiological system of information processing.Journal: Communications biologyIn common: scikit-learn, pandas, SciPy, 2 other tools, 2 references
- [7] doi:10.1038/s41467-026-74565-0 [code]
- The functional neurobiology of dispositions towards negative emotions.Journal: Nature communicationsIn common: lme4, ggplot2, tidyverse, 3 other tools, 1 reference
- [8] doi:10.1073/pnas.2217232120 [code]
- Brain imaging and neuropsychological assessment of individuals recovered from a mild to moderate SARS-CoV-2 infection.Journal: Proceedings of the National Academy of Sciences of the United States of AmericaIn common: ggplot2, tidyverse, scikit-learn, 4 other tools, 1 reference
- [9] doi:10.1017/pen.2022.3 [code]
- Perfectionism-related variations in error processing in a task with increased response selection complexity.Journal: Personality neuroscienceIn common: lmerTest, lme4, ggplot2, 1 other tool, 1 reference
- [10] doi:10.1038/s42003-025-08632-5 [code]
- Human neural organoid microphysiological systems show the building blocks necessary for basic learning and memory.Journal: Communications biologyIn common: scikit-learn, pandas, SciPy, 2 other tools, cellular / molecular, 1 reference
Contribute
The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.
Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.
Claim this paper
Correct its record
Say what each link of this record is, remove the ones that are not the paper's, add the ones that are missing. The correction becomes a new version of the record, in its Versions section.
Validate its tracing map
You validate the map as this page shows it: 1 repository of the authors' code, each at its verified commit and with its license, 16 scripts, and 15 matches between paragraphs and code (see the Code and Map sections). It then receives a DOI on Zenodo, with you (your ORCID iD) and OSCR as its creators; the code itself is not deposited.
The map's fingerprint: sha256:a4fd0ae8cb7d6209…
Add the badge to the code's README
If you are an author or a maintainer of this paper's code, you can add this badge. It says what OSCR knows of the paper (its code, its validated tracing map) and links the code to this page. Paste one of these snippets under the README's title. Keep the two marks around it, so that the badge can be replaced in place and is never added twice. OSCR changes nothing in your repository.
Markdown
<!-- oscr-badge:start -->
[](https://openscicode.org/badge/doi/10.1177/0271678x261485404)
<!-- oscr-badge:end -->reStructuredText
.. oscr-badge:start
.. image:: https://openscicode.org/badge/doi/10.1177/0271678x261485404.svg
:target: https://openscicode.org/badge/doi/10.1177/0271678x261485404
:alt: OSCR
.. oscr-badge:endHTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.1177/0271678x261485404"><img src="https://openscicode.org/badge/doi/10.1177/0271678x261485404.svg" alt="OSCR"></a>
<!-- oscr-badge:end -->On GitHub, open the README, choose Edit (the pencil), paste the snippet under the title, then "Commit changes".
If you choose to, you can have the badge kept up to date for you, with oscr badge --write from the command line or with the GitHub Action in your repository's workflows. Every other form and option (a card, AsciiDoc, LaTeX, shields.io) is on the badge page.
Request its removal
To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).
Discussion, reproductions, activity
Discussion: code errors and mismatches between the code and the paper about this paper and its code, from readers and its authors, beside the paper and its tracing map.
Report a mismatch between the code and the paper · Report a code error (signed in)
Reproductions: reports from readers who ran the authors' code and did not get the paper's result, what they ran, with which environment, commit and data.
Report a reproduction failure (signed in)
Versions of its code: the releases of its code tied to a version of this paper (the preprint, the accepted manuscript, the version of record…), each with the tracing map versioned with it.
Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.
