Gene-distance correlation — astrocytes and oligodendrocytes

Published

March 26, 2026

Here we extend the same gene-distance correlation approach to astrocytes and oligodendrocytes to test whether Layers 2 and 3 also show plaque-proximity gradients:

Layer 2 — Amplification (Astrocytes): Ifitm3, Ccl2, Ccl12, Cxcl10, Tnf, Agt
Layer 3 — Remodeling (Oligodendrocytes): Abca1, Abcg1, Apod, Srebf1, Got1↓, Ldhb↓

Setup and data loading

Code

source("R/setup.R")
library(ggrepel)
library(patchwork)
library(ggVennDiagram)

data <- load_slides_and_distances("full")
slide1 <- data$slide1; slide2 <- data$slide2
cell_distances_all <- data$distances_all


## Identify cell-types of interest

ct_astro <- "Astrocyte"
ct_oligo <- "Oligodendrocyte"

Subset objects and distances

Code

# Subset Seurat objects to target cell type
subset_by_type <- function(slide_list, cell_types) {
  map(slide_list, \(obj) {
    keep <- obj$annotatedclusters %in% cell_types
    subset(obj, cells = colnames(obj)[keep])
  })
}

slides_astro <- subset_by_type(list(slide1, slide2), ct_astro)
slides_oligo <- subset_by_type(list(slide1, slide2), ct_oligo)

# Subset distance table to matching cells
dist_astro <- cell_distances_all |>
  filter(cell %in% unlist(map(slides_astro, colnames)))

dist_oligo <- cell_distances_all |>
  filter(cell %in% unlist(map(slides_oligo, colnames)))

message(sprintf("Astrocytes: %d cells", nrow(dist_astro)))
message(sprintf("Oligodendrocytes: %d cells", nrow(dist_oligo)))

Cell counts per group

Code

build_count_table <- function(dist_df, label) {
  dist_df |>
    mutate(
      prox_caa   = if_else(dist_caa < 50,   "proximal", "distal"),
      prox_paren = if_else(dist_paren < 50,  "proximal", "distal")
    ) |>
    group_by(sample_ID, Treatment.Group) |>
    summarise(
      n_total       = n(),
      n_prox_caa    = sum(prox_caa == "proximal"),
      n_prox_paren  = sum(prox_paren == "proximal"),
      .groups = "drop"
    ) |>
    mutate(cell_type = label)
}

bind_rows(
  build_count_table(dist_astro, "Astrocytes"),
  build_count_table(dist_oligo, "Oligodendrocytes")
) |>
  kbl(caption = "Cell counts per sample and proximity group") |>
  kable_styling("striped", full_width = FALSE)

Cell counts per sample and proximity group
sample_ID	Treatment.Group	n_total	n_prox_caa	n_prox_paren	cell_type
KK4_464	Adu	8288	483	3351	Astrocytes
KK4_465	IgG	7772	432	4162	Astrocytes
KK4_492	Adu	10252	755	3818	Astrocytes
KK4_496	IgG	8324	647	4573	Astrocytes
KK4_502	Adu	6782	452	3226	Astrocytes
KK4_504	IgG	7071	572	3123	Astrocytes
KK4_464	Adu	11692	239	5017	Oligodendrocytes
KK4_465	IgG	11446	212	5993	Oligodendrocytes
KK4_492	Adu	13727	486	5645	Oligodendrocytes
KK4_496	IgG	12122	449	6549	Oligodendrocytes
KK4_502	Adu	11097	351	5464	Oligodendrocytes
KK4_504	IgG	10240	405	4982	Oligodendrocytes

Gene-distance correlation functions

compute_gene_distance_correlation() is centralised in R/setup.R (sourced in setup chunk). Cell-type-specific analyses use min_expressing_cells = 50 (lower than the default of 100 to accommodate smaller cell populations).

Code

plot_treatment_scatter <- function(cor_adu, cor_igg, title) {
  df <- inner_join(
    cor_adu |> select(gene, rho_flip_adu = rho_flip, sig_adu = sig),
    cor_igg |> select(gene, rho_flip_igg = rho_flip, sig_igg = sig),
    by = "gene"
  ) |>
    mutate(
      concordance = case_when(
        sig_adu == "enriched_near" & sig_igg == "enriched_near" ~ "both near",
        sig_adu == "enriched_far"  & sig_igg == "enriched_far"  ~ "both far",
        sig_adu != "ns"            & sig_igg != "ns"            ~ "discordant",
        sig_adu != "ns"            | sig_igg != "ns"            ~ "one sig",
        TRUE                                                     ~ "ns"
      )
    )

  colour_map <- c(
    "both near"  = "#E07B39",
    "both far"   = "#7B68AE",
    "discordant" = "#C0392B",
    "one sig"    = "#5DADE2",
    "ns"         = "grey80"
  )

  label_df <- df |> filter(concordance %in% c("both near", "both far", "discordant"))

  ggplot(df, aes(x = rho_flip_igg, y = rho_flip_adu)) +
    geom_point(aes(colour = concordance), size = 1.8, alpha = 0.6) +
    geom_abline(slope = 1, intercept = 0, linetype = "dashed", colour = "grey40") +
    geom_hline(yintercept = 0, colour = "grey70", linewidth = 0.3) +
    geom_vline(xintercept = 0, colour = "grey70", linewidth = 0.3) +
    geom_text_repel(
      data = label_df, aes(label = gene, colour = concordance),
      size = 3.5, fontface = "italic", max.overlaps = 20, show.legend = FALSE
    ) +
    scale_colour_manual(values = colour_map) +
    labs(title = title, x = "rho_flip — IgG", y = "rho_flip — Aducanumab", colour = NULL) +
    theme_minimal(base_size = 13) +
    theme(legend.position = "bottom", panel.grid.minor = element_blank(),
          panel.grid.major = element_line(colour = "grey92"))
}

Code

plot_gene_distance_correlation <- function(
    results,
    title          = "Gene-distance correlation",
    n_top          = 5,
    genes_to_label = NULL
) {
  top_pos    <- results |> filter(sig == "enriched_near") |> slice_min(rank, n = n_top)
  top_neg    <- results |> filter(sig == "enriched_far")  |> slice_max(rank, n = n_top)
  auto_label <- bind_rows(top_pos, top_neg)

  if (!is.null(genes_to_label)) {
    user_genes <- results |> filter(gene %in% genes_to_label)
    auto_label <- bind_rows(auto_label, user_genes) |> distinct(gene, .keep_all = TRUE)
  }

  colour_map <- c(
    "enriched_near" = "#E07B39",
    "enriched_far"  = "#7B68AE",
    "ns"            = "grey70"
  )

  ggplot(results, aes(x = rank, y = rho_flip)) +
    geom_point(aes(colour = sig), size = 2, alpha = 0.5) +
    geom_hline(yintercept = 0, linetype = "dashed", colour = "grey40") +
    geom_text_repel(
      data        = auto_label,
      aes(label = gene, colour = sig),
      size        = 5,
      fontface    = "italic",
      box.padding = 0.4,
      max.overlaps = 20,
      show.legend = FALSE,
      seed = 23
    ) +
    scale_colour_manual(
      values = colour_map,
      labels = c(
        "enriched_near" = "Enriched near plaque (padj < 0.05)",
        "enriched_far"  = "Enriched far from plaque (padj < 0.05)",
        "ns"            = "Not significant"
      )
    ) +
    labs(
      title  = title,
      x      = "Gene rank",
      y      = "Spearman rho (flipped)",
      colour = NULL
    ) +
    theme_minimal(base_size = 14) +
    theme(
      legend.position    = "bottom",
      panel.grid.minor   = element_blank(),
      panel.grid.major.x = element_blank()
    )
}

Astrocyte gene-distance correlations

Candidate genes from DEG_consolidated model (Layer 2 — Amplification): - Near plaque: Ifitm3, Ccl2, Ccl12, Cxcl10, Agt, Gfap, Vim, S100b - Far from plaque: homeostatic astrocyte markers

Code

astro_candidate_genes <- c(
  "Ifitm3", "Ccl2", "Ccl12", "Cxcl10", "Agt",
  "Gfap", "Vim", "S100b", "Apod", "Abca1", "Plin4"
)

cor_caa_astro   <- compute_gene_distance_correlation(
  slides_astro, dist_astro, distance_col = "dist_caa",   min_expressing_cells = 50)
cor_paren_astro <- compute_gene_distance_correlation(
  slides_astro, dist_astro, distance_col = "dist_paren", min_expressing_cells = 50)
cor_any_astro   <- compute_gene_distance_correlation(
  slides_astro, dist_astro, distance_col = "dist_any",   min_expressing_cells = 50)

# qs2::qs_save(cor_caa_astro,   "spatial/astro_cor_caa.qs2")
# qs2::qs_save(cor_paren_astro, "spatial/astro_cor_paren.qs2")
# qs2::qs_save(cor_any_astro,   "spatial/astro_cor_any.qs2")

Code

dist_astro_adu <- dist_astro |> filter(Treatment.Group == "Adu")
dist_astro_igg <- dist_astro |> filter(Treatment.Group == "IgG")

for (trt in c("adu", "igg")) {
  dist_trt <- if (trt == "adu") dist_astro_adu else dist_astro_igg
  for (dcol in c("dist_caa", "dist_paren", "dist_any")) {
    res    <- compute_gene_distance_correlation(slides_astro, dist_trt,
                distance_col = dcol, min_expressing_cells = 50)
    suffix <- sub("dist_", "", dcol)
    qs2::qs_save(res, paste0("spatial/astro_cor_", suffix, "_", trt, ".qs2"))
  }
}

Code

astro_candidate_genes <- c(
  "Ifitm3", "Ccl2", "Ccl12", "Cxcl10", "Agt",
  "Gfap", "Vim", "S100b", "Apod", "Abca1", "Plin4"
)

cor_caa_astro   <- qs2::qs_read("spatial/astro_cor_caa.qs2")
cor_paren_astro <- qs2::qs_read("spatial/astro_cor_paren.qs2")
cor_any_astro   <- qs2::qs_read("spatial/astro_cor_any.qs2")

astro_cor_caa_adu   <- qs2::qs_read("spatial/astro_cor_caa_adu.qs2")
astro_cor_caa_igg   <- qs2::qs_read("spatial/astro_cor_caa_igg.qs2")
astro_cor_paren_adu <- qs2::qs_read("spatial/astro_cor_paren_adu.qs2")
astro_cor_paren_igg <- qs2::qs_read("spatial/astro_cor_paren_igg.qs2")
astro_cor_any_adu   <- qs2::qs_read("spatial/astro_cor_any_adu.qs2")
astro_cor_any_igg   <- qs2::qs_read("spatial/astro_cor_any_igg.qs2")

Rank plots — astrocytes

Code

p_caa_astro   <- plot_gene_distance_correlation(
  cor_caa_astro,
  title          = "Astrocytes — CAA",
  genes_to_label = astro_candidate_genes
)

p_paren_astro <- plot_gene_distance_correlation(
  cor_paren_astro,
  title          = "Astrocytes — Parenchymal",
  genes_to_label = astro_candidate_genes
)

(p_caa_astro | p_paren_astro) +
  plot_annotation(title = "Astrocyte gene-distance correlation") +
  plot_layout(guides = "collect") &
  theme(legend.position = "bottom")

Spearman gene-distance correlation for astrocytes. Positive rho_flip = enriched near plaque. Top 5 enriched-near and enriched-far genes are automatically labeled; candidate Layer 2 genes (Ifitm3, Ccl2, Ccl12, Cxcl10, Agt, Gfap, Vim) are additionally labeled if significant.

Top astrocyte genes

Code

show_top <- function(results, label, n = 10) {
  bind_rows(
    results |> filter(sig == "enriched_near") |> slice_min(rank, n = n),
    results |> filter(sig == "enriched_far")  |> slice_max(rank, n = n)
  ) |>
    select(gene, rho, rho_flip, pval, padj, rank, sig) |>
    mutate(
      across(c(rho, rho_flip), \(x) round(x, 4)),
      across(c(pval, padj), \(x) ifelse(x < 1e-300, "< 1e-300", formatC(x, format = "e", digits = 2)))
    ) |>
    kbl(caption = label) |>
    kable_styling("striped", full_width = FALSE)
}

show_top(cor_caa_astro,   "Top genes — astrocytes, distance to CAA")

Top genes — astrocytes, distance to CAA
gene	rho	rho_flip	pval	padj	rank	sig
Hmgcs1	-0.1234	0.1234	1.04e-163	1.59e-161	1	enriched_near
Glul	-0.1183	0.1183	1.08e-150	9.89e-149	2	enriched_near
Slc17a7	-0.1074	0.1074	2.37e-124	1.35e-122	3	enriched_near
Fdft1	-0.0993	0.0993	1.67e-106	7.64e-105	4	enriched_near
Reln	-0.0928	0.0928	2.86e-93	1.00e-91	5	enriched_near
Msmo1	-0.0926	0.0926	7.58e-93	2.48e-91	6	enriched_near
Mrc1	-0.0920	0.0920	1.17e-91	3.57e-90	7	enriched_near
Vtn	-0.0846	0.0846	8.46e-78	2.15e-76	8	enriched_near
Gfap	-0.0843	0.0843	3.74e-77	9.00e-76	9	enriched_near
Cux2	-0.0798	0.0798	2.44e-69	5.07e-68	10	enriched_near
Mbp	0.1661	-0.1661	4.82e-297	2.20e-294	457	enriched_far
Cst3	0.1302	-0.1302	3.43e-182	7.84e-180	456	enriched_far
Abca1	0.1223	-0.1223	7.72e-161	8.82e-159	455	enriched_far
Agt	0.1137	-0.1137	3.55e-139	2.70e-137	454	enriched_far
Cd81	0.1129	-0.1129	3.04e-137	1.98e-135	453	enriched_far
Srebf1	0.1045	-0.1045	9.48e-118	4.81e-116	452	enriched_far
Pgm1	0.0973	-0.0973	2.69e-102	1.12e-100	451	enriched_far
Folh1	0.0941	-0.0941	1.00e-95	3.82e-94	450	enriched_far
Vcam1	0.0873	-0.0873	1.42e-82	4.05e-81	449	enriched_far
Nts	0.0872	-0.0872	1.61e-82	4.33e-81	448	enriched_far

Code

show_top(cor_paren_astro, "Top genes — astrocytes, distance to parenchymal")

Top genes — astrocytes, distance to parenchymal
gene	rho	rho_flip	pval	padj	rank	sig
Serpina3n	-0.3880	0.3880	< 1e-300	< 1e-300	1	enriched_near
Gfap	-0.3184	0.3184	< 1e-300	< 1e-300	2	enriched_near
Ctsd	-0.2286	0.2286	< 1e-300	< 1e-300	3	enriched_near
Mbp	-0.2224	0.2224	< 1e-300	< 1e-300	4	enriched_near
Gpnmb	-0.1947	0.1947	< 1e-300	< 1e-300	5	enriched_near
Ctsb	-0.1910	0.1910	< 1e-300	< 1e-300	6	enriched_near
Tmsb4x	-0.1686	0.1686	< 1e-300	< 1e-300	7	enriched_near
Vim	-0.1675	0.1675	< 1e-300	1.44e-300	8	enriched_near
Lpl	-0.1661	0.1661	4.14e-297	9.96e-296	9	enriched_near
Ftl1	-0.1660	0.1660	1.10e-296	2.52e-295	10	enriched_near
Gja1	0.3079	-0.3079	< 1e-300	< 1e-300	457	enriched_far
Glul	0.2747	-0.2747	< 1e-300	< 1e-300	456	enriched_far
Cbs	0.2328	-0.2328	< 1e-300	< 1e-300	455	enriched_far
Htra1	0.2182	-0.2182	< 1e-300	< 1e-300	454	enriched_far
Phgdh	0.2140	-0.2140	< 1e-300	< 1e-300	453	enriched_far
Glud1	0.2010	-0.2010	< 1e-300	< 1e-300	452	enriched_far
Pygb	0.1965	-0.1965	< 1e-300	< 1e-300	451	enriched_far
Fdft1	0.1763	-0.1763	< 1e-300	< 1e-300	450	enriched_far
Msmo1	0.1747	-0.1747	< 1e-300	< 1e-300	449	enriched_far
Plpp3	0.1703	-0.1703	< 1e-300	< 1e-300	448	enriched_far

Treatment-split rank plots — astrocytes

Code

plot_gene_distance_correlation(astro_cor_caa_adu,
  title = "Astrocytes — CAA, Aducanumab", genes_to_label = astro_candidate_genes) |
plot_gene_distance_correlation(astro_cor_caa_igg,
  title = "Astrocytes — CAA, IgG",        genes_to_label = astro_candidate_genes)

Code

plot_gene_distance_correlation(astro_cor_paren_adu,
  title = "Astrocytes — Parenchymal, Aducanumab", genes_to_label = astro_candidate_genes) |
plot_gene_distance_correlation(astro_cor_paren_igg,
  title = "Astrocytes — Parenchymal, IgG",        genes_to_label = astro_candidate_genes)

Code

plot_treatment_scatter(astro_cor_caa_adu,   astro_cor_caa_igg,   "Astrocytes — CAA, Adu vs IgG") |
plot_treatment_scatter(astro_cor_paren_adu, astro_cor_paren_igg, "Astrocytes — Parenchymal, Adu vs IgG") |
plot_treatment_scatter(astro_cor_any_adu,   astro_cor_any_igg,   "Astrocytes — Any plaque, Adu vs IgG")

Venn diagram — astrocyte enriched-near overlap

Code

near_caa_astro   <- cor_caa_astro   |> filter(padj < 0.05, rho_flip > 0) |> pull(gene)
near_paren_astro <- cor_paren_astro |> filter(padj < 0.05, rho_flip > 0) |> pull(gene)

ggVennDiagram(
  list(CAA = near_caa_astro, Parenchymal = near_paren_astro),
  label_alpha = 0, label = "count", shape_id = "201"
) +
  labs(title = "Astrocytes — enriched-near genes (padj < 0.05)") +
  theme(legend.position = "none") +
  scale_fill_distiller(palette = "RdBu")

Overlap of genes significantly enriched near plaques (padj < 0.05, rho_flip > 0) in astrocytes, between CAA and parenchymal analyses.

Code

cat("Common enriched-near astrocyte genes (CAA and parenchymal):\n",
    paste(intersect(near_caa_astro, near_paren_astro), collapse = ", "))

Common enriched-near astrocyte genes (CAA and parenchymal):
 Gfap, Vim, Cox4i1, Whrn, Ifitm3, Gpnmb, Ifitm2, Lyz2, Uba52, Aqp4, Cd44, Cerk, Cd63, App, Spp1, Cx3cr1, C1qc, Hk2, Pvalb, St3gal6, Ldha, Chst11, Cd14, Flt1, Fxyd5, Ftl1, C1qa, Tshz2, Pdha1, Ogdh, Jund, Bcl2a1b, C3, Tmsb4x, Pkm, Ccl3, Mt2, Cd68, Tagln, B3gat3, Arsb, Trem2, Abca7, Hexb, Dlst, Ocln, Gns, Got2, Lgals3, Cxcl10, Pgd

Oligodendrocyte gene-distance correlations

Candidate genes from DEG_consolidated model (Layer 3 — Remodeling): - Near plaque: Abca1, Abcg1, Srebf1, Apod, Aspa, Mbp, Plp1 - Far from plaque: Got1, Ldhb, Mdh2, Hk1 (metabolic downregulation)

Code

oligo_candidate_genes <- c(
  "Abca1", "Abcg1", "Srebf1", "Apod", "Aspa",
  "Mbp", "Plp1", "Got1", "Ldhb", "Mdh2", "Hk1"
)

cor_caa_oligo   <- compute_gene_distance_correlation(
  slides_oligo, dist_oligo, distance_col = "dist_caa",   min_expressing_cells = 50)
cor_paren_oligo <- compute_gene_distance_correlation(
  slides_oligo, dist_oligo, distance_col = "dist_paren", min_expressing_cells = 50)
cor_any_oligo   <- compute_gene_distance_correlation(
  slides_oligo, dist_oligo, distance_col = "dist_any",   min_expressing_cells = 50)

# qs2::qs_save(cor_caa_oligo,   "spatial/oligo_cor_caa.qs2")
# qs2::qs_save(cor_paren_oligo, "spatial/oligo_cor_paren.qs2")
# qs2::qs_save(cor_any_oligo,   "spatial/oligo_cor_any.qs2")

Code

dist_oligo_adu <- dist_oligo |> filter(Treatment.Group == "Adu")
dist_oligo_igg <- dist_oligo |> filter(Treatment.Group == "IgG")

for (trt in c("adu", "igg")) {
  dist_trt <- if (trt == "adu") dist_oligo_adu else dist_oligo_igg
  for (dcol in c("dist_caa", "dist_paren", "dist_any")) {
    res    <- compute_gene_distance_correlation(slides_oligo, dist_trt,
                distance_col = dcol, min_expressing_cells = 50)
    suffix <- sub("dist_", "", dcol)
    qs2::qs_save(res, paste0("spatial/oligo_cor_", suffix, "_", trt, ".qs2"))
  }
}

Code

oligo_candidate_genes <- c(
  "Abca1", "Abcg1", "Srebf1", "Apod", "Aspa",
  "Mbp", "Plp1", "Got1", "Ldhb", "Mdh2", "Hk1"
)

cor_caa_oligo   <- qs2::qs_read("spatial/oligo_cor_caa.qs2")
cor_paren_oligo <- qs2::qs_read("spatial/oligo_cor_paren.qs2")
cor_any_oligo   <- qs2::qs_read("spatial/oligo_cor_any.qs2")

oligo_cor_caa_adu   <- qs2::qs_read("spatial/oligo_cor_caa_adu.qs2")
oligo_cor_caa_igg   <- qs2::qs_read("spatial/oligo_cor_caa_igg.qs2")
oligo_cor_paren_adu <- qs2::qs_read("spatial/oligo_cor_paren_adu.qs2")
oligo_cor_paren_igg <- qs2::qs_read("spatial/oligo_cor_paren_igg.qs2")
oligo_cor_any_adu   <- qs2::qs_read("spatial/oligo_cor_any_adu.qs2")
oligo_cor_any_igg   <- qs2::qs_read("spatial/oligo_cor_any_igg.qs2")

Rank plots — oligodendrocytes

Code

p_caa_oligo   <- plot_gene_distance_correlation(
  cor_caa_oligo,
  title          = "Oligodendrocytes — CAA",
  genes_to_label = oligo_candidate_genes
)

p_paren_oligo <- plot_gene_distance_correlation(
  cor_paren_oligo,
  title          = "Oligodendrocytes — Parenchymal",
  genes_to_label = oligo_candidate_genes
)

(p_caa_oligo | p_paren_oligo) +
  plot_annotation(title = "Oligodendrocyte gene-distance correlation") +
  plot_layout(guides = "collect") &
  theme(legend.position = "bottom")

Spearman gene-distance correlation for oligodendrocytes. Top 5 enriched-near and enriched-far genes are automatically labeled; candidate Layer 3 genes (Abca1, Abcg1, Srebf1, Apod, Got1, Ldhb) are additionally labeled if present.

Top oligodendrocyte genes

Code

show_top(cor_caa_oligo,   "Top genes — oligodendrocytes, distance to CAA")

Top genes — oligodendrocytes, distance to CAA
gene	rho	rho_flip	pval	padj	rank	sig
Ttr	-0.0897	0.0897	1.25e-125	2.87e-123	1	enriched_near
Slc17a7	-0.0654	0.0654	1.74e-67	1.59e-65	2	enriched_near
Ptgds	-0.0513	0.0513	3.17e-42	1.62e-40	3	enriched_near
Opcml	-0.0504	0.0504	9.37e-41	4.29e-39	4	enriched_near
Grin2a	-0.0488	0.0488	2.24e-38	8.53e-37	5	enriched_near
Lamp5	-0.0473	0.0473	4.53e-36	1.60e-34	6	enriched_near
Slc17a6	-0.0437	0.0437	3.86e-31	1.18e-29	7	enriched_near
Snap25	-0.0404	0.0404	9.29e-27	2.50e-25	8	enriched_near
Atp1b2	-0.0387	0.0387	1.05e-24	2.41e-23	9	enriched_near
Whrn	-0.0386	0.0386	1.24e-24	2.70e-23	10	enriched_near
Apod	0.0913	-0.0913	4.27e-130	1.96e-127	458	enriched_far
Tac1	0.0833	-0.0833	1.81e-108	2.76e-106	457	enriched_far
Nts	0.0811	-0.0811	6.86e-103	7.85e-101	456	enriched_far
Mbp	0.0582	-0.0582	8.17e-54	6.24e-52	455	enriched_far
Ermn	0.0575	-0.0575	1.30e-52	8.48e-51	454	enriched_far
Tmsb4x	0.0538	-0.0538	3.67e-46	2.10e-44	453	enriched_far
Prdx1	0.0494	-0.0494	3.02e-39	1.26e-37	452	enriched_far
Serpina3n	0.0461	-0.0461	1.94e-34	6.34e-33	451	enriched_far
Ptges3	0.0431	-0.0431	3.02e-30	8.64e-29	450	enriched_far
Cldn11	0.0400	-0.0400	2.63e-26	6.70e-25	449	enriched_far

Code

show_top(cor_paren_oligo, "Top genes — oligodendrocytes, distance to parenchymal")

Top genes — oligodendrocytes, distance to parenchymal
gene	rho	rho_flip	pval	padj	rank	sig
Cst7	-0.2317	0.2317	< 1e-300	< 1e-300	1	enriched_near
Fabp5	-0.1570	0.1570	< 1e-300	< 1e-300	2	enriched_near
Ogdhl	-0.1453	0.1453	< 1e-300	< 1e-300	3	enriched_near
Ctsb	-0.1429	0.1429	< 1e-300	< 1e-300	4	enriched_near
Serpina3n	-0.1427	0.1427	< 1e-300	< 1e-300	5	enriched_near
Ctsd	-0.1413	0.1413	< 1e-300	< 1e-300	6	enriched_near
Gpnmb	-0.1410	0.1410	< 1e-300	< 1e-300	7	enriched_near
App	-0.1311	0.1311	6.09e-267	3.48e-265	8	enriched_near
Cd68	-0.1204	0.1204	1.80e-225	8.24e-224	9	enriched_near
Bcl2a1b	-0.1178	0.1178	8.35e-216	3.48e-214	10	enriched_near
Ubc	0.1279	-0.1279	4.30e-254	2.19e-252	458	enriched_far
Phgdh	0.1140	-0.1140	4.66e-202	1.78e-200	457	enriched_far
Pygb	0.0921	-0.0921	2.42e-132	4.81e-131	456	enriched_far
Cbs	0.0792	-0.0792	2.86e-98	3.97e-97	455	enriched_far
Aldh2	0.0742	-0.0742	1.69e-86	2.03e-85	454	enriched_far
Gja1	0.0724	-0.0724	2.23e-82	2.49e-81	453	enriched_far
Clu	0.0717	-0.0717	9.27e-81	1.01e-79	452	enriched_far
Ttr	0.0714	-0.0714	3.35e-80	3.57e-79	451	enriched_far
Psap	0.0703	-0.0703	9.83e-78	9.79e-77	450	enriched_far
Chst1	0.0688	-0.0688	1.52e-74	1.48e-73	449	enriched_far

Treatment-split rank plots — oligodendrocytes

Code

plot_gene_distance_correlation(oligo_cor_caa_adu,
  title = "Oligodendrocytes — CAA, Aducanumab", genes_to_label = oligo_candidate_genes) |
plot_gene_distance_correlation(oligo_cor_caa_igg,
  title = "Oligodendrocytes — CAA, IgG",        genes_to_label = oligo_candidate_genes)

Code

plot_gene_distance_correlation(oligo_cor_paren_adu,
  title = "Oligodendrocytes — Parenchymal, Aducanumab", genes_to_label = oligo_candidate_genes) |
plot_gene_distance_correlation(oligo_cor_paren_igg,
  title = "Oligodendrocytes — Parenchymal, IgG",        genes_to_label = oligo_candidate_genes)

Code

plot_treatment_scatter(oligo_cor_caa_adu,   oligo_cor_caa_igg,   "Oligodendrocytes — CAA, Adu vs IgG") |
plot_treatment_scatter(oligo_cor_paren_adu, oligo_cor_paren_igg, "Oligodendrocytes — Parenchymal, Adu vs IgG") |
plot_treatment_scatter(oligo_cor_any_adu,   oligo_cor_any_igg,   "Oligodendrocytes — Any plaque, Adu vs IgG")

Venn diagram — oligodendrocyte enriched-near overlap

Code

near_caa_oligo   <- cor_caa_oligo   |> filter(padj < 0.05, rho_flip > 0) |> pull(gene)
near_paren_oligo <- cor_paren_oligo |> filter(padj < 0.05, rho_flip > 0) |> pull(gene)

ggVennDiagram(
  list(CAA = near_caa_oligo, Parenchymal = near_paren_oligo),
  label_alpha = 0, label = "count", shape_id = "201"
) +
  labs(title = "Oligodendrocytes — enriched-near genes (padj < 0.05)") +
  theme(legend.position = "none") +
  scale_fill_distiller(palette = "RdBu")

Overlap of genes significantly enriched near plaques (padj < 0.05, rho_flip > 0) in oligodendrocytes.

Code

cat("Common enriched-near oligodendrocyte genes (CAA and parenchymal):\n",
    paste(intersect(near_caa_oligo, near_paren_oligo), collapse = ", "))

Common enriched-near oligodendrocyte genes (CAA and parenchymal):
 Slc17a7, Ptgds, Opcml, Slc17a6, Snap25, Extl3, Bhlhe40, Gls, B3gat3, Pvalb, Acly, Pkm, Timp3, Arsb, Cerk, Got1, Ldhb, Mdh1, Cux2, Oxct1, Cox4i1, Glul, Got2, Ogdhl, Hbb-bs, Akt1, Pank3, Chst11, Hk1, Flt1, Cd74, Gns, Tshz2, Uqcrc1, Plcb4, H2-Eb1, Acsl4, Rorb, Hsd11b1, Pdhb, Pmm1, Sgpl1, Gpnmb, Cldn5, Abhd17a, Fezf2, Pdha1, Smpd1, Maf, Sorl1

Cross-cell-type comparison

Astrocytes vs oligodendrocytes — candidate gene heatmap

Visualise the rho_flip values for all DEG_consolidated candidate genes across both cell types and both plaque types in a single heatmap.

Code

all_candidates <- union(astro_candidate_genes, oligo_candidate_genes)

heatmap_df <- bind_rows(
  cor_caa_astro   |> mutate(cell_type = "Astrocytes",      plaque = "CAA"),
  cor_paren_astro |> mutate(cell_type = "Astrocytes",      plaque = "Parenchymal"),
  cor_caa_oligo   |> mutate(cell_type = "Oligodendrocytes", plaque = "CAA"),
  cor_paren_oligo |> mutate(cell_type = "Oligodendrocytes", plaque = "Parenchymal")
) |>
  filter(gene %in% all_candidates) |>
  mutate(
    facet = paste(cell_type, plaque, sep = " — "),
    sig_label = case_when(
      padj < 0.001 ~ "***",
      padj < 0.01  ~ "**",
      padj < 0.05  ~ "*",
      TRUE         ~ ""
    )
  )

ggplot(heatmap_df, aes(x = facet, y = gene, fill = rho_flip)) +
  geom_tile(colour = "white", linewidth = 0.5) +
  geom_text(aes(label = sig_label), size = 4, vjust = 0.75) +
  scale_fill_gradient2(
    low  = "#7B68AE", mid = "white", high = "#E07B39",
    midpoint = 0,
    name = "rho_flip\n(+ = near plaque)"
  ) +
  labs(
    title = "Candidate gene-distance correlations",
    subtitle = "* padj < 0.05  ** padj < 0.01  *** padj < 0.001",
    x = NULL, y = NULL
  ) +
  theme_minimal(base_size = 13) +
  theme(axis.text.x = element_text(angle = 30, hjust = 1))

Spearman rho_flip for candidate Layer 2 (astrocyte) and Layer 3 (oligodendrocyte) genes across CAA and parenchymal distance. Positive (orange) = enriched near plaque; negative (purple) = enriched far from plaque. Grey fill = gene not tested (< 50 expressing cells).