Talin Loop Paper Data Analysis

Author

Jannik Buhr

Published

April 11, 2022

Tip

To reproduce the graphics generated here, download the raw data from the archive and replace the value of the target data_path_prefix in _targets.R with the path to that data. Then use the next two code chunks to load packages and functions and run the targets pipeline (tar_make()). The raw data will then be processed such that you can load it in this document.

Setup

Code 
source("./R/globals.R")

Data processing pipeline

Code 
tar_visnetwork(callr_arguments = list(show = FALSE))

Plots

Asp125 and Glu126 missing -> sequence shifted by 2 afterwards.

Load data

Code 
f0f1_run1 <- tar_read(f0f1_rot_sample_distances_per_run, branches = 1:60)
tar_load(CUTOFF)
tar_load(from_first_contact)
tar_load(interacting)
tar_load(sequence)
tar_load(rolling)
tar_load(stuck)
tar_load(ferm_interacting)
tar_load(ferm_annotation)
tar_load(ferm_colors)
tar_load(bool_colors)
tar_load(f0f1_pulling_smooth)
tar_load(f0f1_pulling_peaks)
tar_load(ferm_pip_sites)
tar_load(f0f1_pulling_interacting)
tar_load(ferm_rmsf)
tar_load(nmr_rmsf)

aa <- read_tsv("./assets/tables/aa.tsv") |>
  mutate(
    charge = str_remove(`Net charge at pH 7.4`, ",.+")
  ) |>
  select(`1`, `3`, charge)

CHARGE_COLORS = c(
  "Negative" = HITS_MAGENTA,
  "Positive" = HITS_BLUE,
  "Neutral" = "grey50"
)

Rotational sampling of F0F1

  • Data loss due to disk failure:
    • run4/angle_348: corrupted
    • run2/angle_204: corrupted only frames 00143 and 00166
    • run3/angle_234: corrupted after first 4 frames
Code 
anti_join(
  expand_grid(run = 1:6, angle = seq(0, 354, 6)),
  distinct(interacting, run, angle)
)
# A tibble: 1 × 2
    run angle
  <int> <dbl>
1     4   348
Code 
interacting |> 
  drop_na() |> 
  count(run, angle) |> 
  filter(n != max(n))
# A tibble: 2 × 3
    run angle     n
  <int> <int> <int>
1     2   204 39203
2     3   234   788
Code 
anti_join(
  expand_grid(frame = 0:200, i = 1:197),
  interacting |> 
    filter(run == 2, angle == 204) |> 
    distinct(frame, i)
  )
# A tibble: 394 × 2
   frame     i
   <int> <int>
 1   143     1
 2   143     2
 3   143     3
 4   143     4
 5   143     5
 6   143     6
 7   143     7
 8   143     8
 9   143     9
10   143    10
# … with 384 more rows
Code 
interacting |> 
  drop_na() |> 
  group_by(run, angle) |> 
  summarise(frame = max(frame)) |> 
  ungroup() |> 
  filter(frame != max(frame))
# A tibble: 1 × 3
    run angle frame
  <int> <int> <int>
1     3   234     3
Code 
paths <- c("./assets/blender/render-rotsample/frame0001.png",
    "./assets/blender/render-rotsample/frame0004.png",
    "./assets/blender/render-rotsample/frame0003.png",
    "./assets/blender/render-rotsample/frame0002.png",
    "./assets/blender/render-rotsample/frame0001.png") |>map_chr(glue)

render_png <- function(p) grid::rasterGrob(png::readPNG(p))
pngs <- map(paths, render_png)

plt <- from_first_contact |>
  group_by(i, angle) |>
  summarise(
      n_pip = mean(n_pip)
      ) |>
  ungroup() |>
  complete(i = 1:197, angle = seq(0, 354, 6), fill = list(n_pip = 0)) |>
  ggplot(aes(i, angle, fill = n_pip)) + 
  geom_raster() +
  scale_fill_gradient(low = "white", high = HITS_BLUE) +
  scale_y_continuous(breaks = seq(0, 360, 90)) +
  coord_cartesian(clip = "off", xlim = c(0, 207)) +
  labs(
      x = "residue index",
      y = "angle [deg]",
      fill = expression(bar(n)[pip2])
      )

grobs <- ggplotGrob(plt)
plt <- plt +
guides(
    fill = guide_colorbar(
      barheight = unit(0.7, "grobheight", data = grobs$grobs),
      frame.colour = "black")
)

anot <- ggplot(ferm_annotation[1:3, ]) +
  aes(
    ymin = 0, ymax = 10,
    xmin = imin, xmax = imax,
    fill = domain) +
  geom_rect(show.legend = FALSE) +
  geom_text(aes(y = 5, x = (imin + imax) / 2, label = domain),
            color = "black") +
  theme_void() +
  coord_cartesian(clip = "off", xlim = c(0, 207)) +
  scale_fill_discrete(type = ferm_colors)

vmd <- wrap_plots(pngs, ncol = 1)

main <- (plt / anot) +
  plot_layout(heights = c(20, 1)) +
  plot_annotation()

((vmd / plot_spacer()) | main) +
  plot_layout(widths = c(1, 10))

Code 
update <- function(data, row, colname, value) {
  col <- rlang::as_name(rlang::enquo(colname))
  data[row, col] <- value
  data
}
Code 
plt <- from_first_contact |>
  group_by(i) |> 
  summarise(n_pip = mean(n_pip)) |>
  left_join(sequence, by = "i") |>
  left_join(aa, by = c('res' = '1')) |>
  update(1, charge, "Positive") |>
  ggplot(aes(x = i, y = n_pip)) + 
  annotate(geom = "rect",
           xmin = 140, xmax = 164,
           ymin = 0, ymax = Inf,
           color = NA,
           fill = "black",
           alpha = 0.05) +
  geom_col(aes(fill = charge),
           alpha = 1,
           width = 0.8,
           show.legend = TRUE) +
  geom_text(aes(label = res, color = charge),
            fontface = "bold",
            vjust = -0.2,
            data = ~ filter(.x, n_pip > 0.1), show.legend = FALSE) +
  scale_fill_manual(
    values = CHARGE_COLORS, aesthetics = c("fill", "color"),
    name = "Net Charge at pH 7.4"
  ) +
  # scale_y_continuous(breaks = seq(0, 360, 90), expand = expansion(mult = c(0, 0.1))) +
  labs(y = expression(n[PIP[2]])) +
  coord_cartesian(clip = "off", xlim = c(0, 207)) +
  labs(
    x = "residue index",
    y = expression(bar(n)[PIP[2]]),
    ) +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    legend.position = c(0, 1),
    legend.justification = c(0, 1),
    legend.title = element_text(size = 12)
  )

grobs <- ggplotGrob(plt)

anot <- ggplot(ferm_annotation[1:3, ]) +
  aes(
    ymin = 0, ymax = 10,
    xmin = imin, xmax = imax,
    fill = domain) +
  geom_rect(show.legend = FALSE) +
  geom_text(aes(y = 5, x = (imin + imax) / 2, label = domain),
            color = "black") +
  theme_void() +
  coord_cartesian(clip = "off", xlim = c(0, 207)) +
  scale_fill_discrete(type = ferm_colors)

(plt / anot) +
  plot_layout(heights = c(20, 1)) +
  plot_annotation()

Code 
plt <- interacting |>
  group_by(frame, angle) |>
  summarise(pip = mean(n_pip)) |>
  ggplot(aes(angle, frame, fill = pip)) +
  geom_raster() +
  scale_fill_gradient(low = "white", high = HITS_BLUE) +
  scale_x_continuous(breaks = seq(0, 360, 90)) +
  guides(color = guide_colorbar(barheight = 16)) +
  labs(
    fill = expression(bar(n)[PIP[2]]),
    y = "time [ns]",
    x = "angle [deg]"
  )
  
grobs <- ggplotGrob(plt)

plt + guides(fill = guide_colorbar(barheight = unit(0.7, "grobheight",
  data = grobs$grobs),
  frame.colour = "black"))

Code 
interacting |>
  group_by(run, angle, i) |>
  arrange(frame) |>
  filter(n_pip > 0) |>
  slice_min(frame) |>
  group_by(i, angle) |>
  summarise(frame = mean(frame)) |>
  ungroup() |>
  mutate(
    region = case_when(
     between(i, 135, 168) ~ " ",
     between(i, 1, 84) ~ "f0",
     between(i, 85, 207) ~ "f1",
     TRUE ~ "rest"
    ),
  ) |>
  ggplot(aes(region, frame)) +
  geom_violin(aes(fill = region)) +
  # ggbeeswarm::geom_quasirandom(width = 0.4) +
  stat_summary(fun = "mean", geom = "crossbar", width = 0.7, size = 1) +
  scale_x_discrete(labels = c("f1-loop", "f0", "f1"), name = "") +
  scale_fill_discrete(type = ferm_colors) +
  guides(fill = "none") +
  labs(y = "first contact time [ns]")

Code 
interacting |>
  group_by(run, angle, i) |>
  arrange(frame) |>
  filter(n_pip > 0) |>
  slice_min(frame) |>
  group_by(i, angle) |>
  summarise(frame = mean(frame)) |>
  ungroup() |>
  mutate(
    region = case_when(
     between(i, 135, 168) ~ "f1-loop",
     between(i, 1, 84) ~ "f0",
     between(i, 85, 207) ~ "f1",
     TRUE ~ "rest"
    ),
  ) |>
  (\(x) t.test(frame ~ region == "f1-loop", data = x))()

    Welch Two Sample t-test

data:  frame by region == "f1-loop"
t = 6.8669, df = 2112.6, p-value = 8.597e-12
alternative hypothesis: true difference in means between group FALSE and group TRUE is not equal to 0
95 percent confidence interval:
  8.634507 15.537756
sample estimates:
mean in group FALSE  mean in group TRUE 
          110.06539            97.97926
Code 
interacting |>
  group_by(run, angle, i) |>
  arrange(frame) |>
  filter(n_pip > 0) |>
  slice_min(frame) |>
  ungroup() |>
  drop_na() |>
  mutate(
    region = case_when(
     between(i, 135, 168) ~ " ",
     between(i, 1, 84) ~ "f0",
     between(i, 85, 207) ~ "f1",
     TRUE ~ "rest"
    ),
    angle = cut(angle, breaks = seq(0, 360, by = 60), include.lowest = TRUE)
  ) |>
  ggplot(aes(angle, fill = region)) +
  geom_bar(position = "dodge2") +
  scale_fill_discrete(type = ferm_colors, labels = c("f1-loop", "f0", "f1")) +
  scale_y_continuous(expand = c(0, 1)) +
  theme(axis.text.x = element_text(size = 10, angle = 45, hjust = 1)) +
  labs(y = "# region made first contact")

Code 
f0f1_stickiness <- rolling |>
  group_by(run, angle) |>
  slice_max(time) |>
  summarise(stuck = n_i > 0) |>
  ungroup()

f0f1_time_ri <- interacting |>
  group_by(run, angle, frame) |>
  summarise(n_i = sum(n_pip > 0)) |>
  ungroup() |>
  left_join(f0f1_stickiness)

plt1 <- f0f1_time_ri |>
  filter(!is.na(stuck)) |>
  mutate(stuck = fct_rev(factor(stuck))) |>
  ggplot(aes(frame, n_i, group = paste(run, angle), color = stuck)) +
  geom_line(aes(alpha = stuck)) +
  geom_smooth(aes(group = stuck, color = stuck), se = FALSE,
              data = ~filter(.x, stuck == TRUE)) +
  scale_color_manual(values = bool_colors, name = "retention") +
  scale_alpha_manual(values = c(`TRUE` = 0.08, `FALSE` = 0.6)) +
  labs(x = "time [ns]",
       y = "# interacting residues") +
  coord_cartesian(clip = "off") +
  guides(color = "none", alpha = "none") +
  theme(
    plot.margin = unit(c(1, 0, 1, 1), "lines")
  )

plt2 <- stuck |>
  ggplot(aes(x = 0, fill = stuck, n, label = paste0(n))) +
  geom_col(position = "stack") +
  geom_text(
    vjust = 1.1,
    position = "stack",
    fontface = "bold",
    size = 3,
    color = "white"
  ) +
  scale_x_continuous(breaks = NULL) +
  scale_fill_manual(values = c("grey30", HITS_MAGENTA, HITS_BLUE)) +
  labs(x = "total", y = "", title = "") +
  theme(
    legend.position = "top",
    axis.text = element_blank(),
    axis.ticks = element_blank(),
    plot.margin = unit(c(1, 1, 0, 0), "lines")
  ) +
  guides(fill = guide_legend(title = ""))


(plt1 | plt2) +
  plot_layout(widths = c(15, 1), guides = 'collect') &
  theme(legend.position = "bottom")

Code 
plt <- f0f1_run1 |>
  filter(r != 1) |>
  ggplot(aes(r, y = stat(density))) +
  geom_density(fill = HITS_BLUE, alpha = 0.6) +
  geom_vline(xintercept = CUTOFF, lty = 2) +
  annotate("text", x = CUTOFF, y = 1,
           label = glue("r = {CUTOFF} nm"), hjust = -0.5,
           fontface = "bold") +
  labs(
    x = "r [nm]",
    y = "#",
    caption = "r ≥ 1 not shown due to cutoff"
  ) +
  coord_cartesian(xlim = c(0, 0.9))

plt

Vertical pulling of F0F1

Code 
plt_contacts <- f0f1_pulling_interacting |>
  group_by(run, frame) |>
  summarise(contacts = sum(n_pip > 0)) |>
  ggplot(aes(frame * 0.03, contacts)) +
  geom_line(aes(group = run, lty = (run == 4)), alpha = 0.6) +
  labs(
    x = "distance [nm]",
    y = "# interacting residues"
  ) +
  guides(lty = "none")

plt_curves <- f0f1_pulling_smooth |> 
  ggplot(aes(time / 1e3 *  0.03, f)) +
  geom_line(aes(group = run, lty = (run == 4)), alpha = 0.3) +
  geom_smooth(aes(group = run, lty = (run == 4)), se = FALSE, color = HITS_BLUE) +
  labs(x = "distance [nm]",
       y = "force [pN]",
       caption = "pulling rate: 0.03 nm/ns") +
  coord_cartesian(ylim = c(0, 150)) +
  guides(lty = "none")

(plt_contacts / plt_curves)

Code 
f0f1_pulling_interacting |> 
  group_by(run) |> 
  filter(frame == max(frame)) |> 
  summarise(sum(n_pip > 0))
# A tibble: 6 × 2
    run `sum(n_pip > 0)`
  <dbl>            <int>
1     1                0
2     2                0
3     3                0
4     4               15
5     5                0
6     6                0
  • Run 4 looks interesting.
  • PIP2s get pulled out of the membrane, one lodges itself near F0, two more get pulled out by the loop.
Code 
# loop: 135 to 168
df <- f0f1_pulling_interacting |>
  arrange(run, frame, i) |>
  mutate(loop = between(i, 135, 168)) |>
  group_by(run, frame, loop) |>
  summarise(contacts = sum(n_pip > 0)) |> 
  group_by(run, loop) |>
  mutate(contacts = zoo::rollmean(contacts, 15, NA)) |>
  pivot_wider(names_from = loop, values_from = contacts) |> 
  ungroup()

head_tail <- function(data, ...) {
  bind_rows(
    slice_max(data, ...),
    slice_min(data, ...)
  )
}

points <- df |>
  filter(!is.na(`FALSE`), !is.na(`TRUE`)) |>
  group_by(run) |>
  slice_head()

arrows_end <- df |>
  filter(!is.na(`FALSE`), !is.na(`TRUE`)) |>
  group_by(run) |>
  head_tail(frame, n = 2) |>
  arrange(run, frame) |>
  mutate(end = rep(1:(n() %/%  2), each = 2)) |>
  filter(end == 2)

df |>
  left_join(
    f0f1_pulling_smooth |> 
      select(run, time, f) |> 
      mutate(frame = time %/% 1e3) |> 
      group_by(run, frame) |> 
      summarise(f = mean(f))
  ) |> 
  ggplot(aes(`TRUE`, `FALSE`, color = f, group = paste(run))) +
  geom_path() +
  geom_point(data = points, color = "black") +
  geom_path(
    aes(group = paste(run, end)),
    color = "black",
    data = arrows_end,
    arrow = arrow(angle = 15, ends = "last", type = "closed",
                  length = unit(0.4, "lines")),
  ) +
  scale_color_viridis_c(direction = -1) +
  coord_equal() +
  labs(
    x = "loop contacts",
    y = "non loop contacts",
    color = "F [pN]"
  )

Code 
colors <- c("#19217d", "#7d127f", "#be0170", "#eb3456", "#ff6d35", "#ffa600")
df |> 
  ggplot(aes(`TRUE`, `FALSE`, color = factor(run), group = paste(run))) +
  geom_path(size = 0.8) +
  geom_point(data = points, color = "black", size = 3) +
  geom_path(
    aes(group = paste(run, end)),
    color = "black",
    data = arrows_end,
    arrow = arrow(angle = 15, ends = "last", type = "closed",
                  length = unit(0.5, "lines")),
  ) +
  # scale_color_gradient(high = "#98cde9", low = "#003063") +
  # scale_colour_brewer(type = "qual", palette = "Set1") +
  scale_color_manual(values = colors) +
  coord_equal() +
  guides(color = "none", alpha = "none") +
  labs(
    x = "loop contacts",
    y = "non loop contacts",
    color = ""
  )

Code 
# rate: 0.03 nm/ns
rate <- 0.03

plt <- f0f1_pulling_interacting |> 
  group_by(frame, i) |> 
  summarise(n_pip = mean(n_pip)) |> 
  mutate(d = frame * rate) |> 
  ggplot(aes(i, d, fill = n_pip)) +
  geom_raster() +
  # facet_wrap(~run) +
  scale_fill_gradient(low = "white", high = HITS_BLUE) +
  coord_cartesian(clip = "off", xlim = c(0, 207)) +
  guides(fill = "none") +
  labs(y = "pulling distance [nm]",
       x = "residue index"
       # caption = "pulling rate: 0.03 nm/ns"/
       )

anot <- ggplot(ferm_annotation[1:3, ]) +
  aes(
    ymin = 0, ymax = 10,
    xmin = imin, xmax = imax,
    fill = domain) +
  geom_rect(show.legend = FALSE) +
  geom_text(aes(y = 5, x = (imin + imax) / 2, label = domain),
            color = "black") +
  theme_void() +
  coord_cartesian(clip = "off", xlim = c(0, 207)) +
  scale_fill_discrete(type = ferm_colors)

(plt / anot) +
  plot_layout(heights = c(20, 1)) +
  plot_annotation()

Equilibrium simulations of FERM

Code 
plt <- ferm_interacting |> 
  filter(run == 6) |> 
  ggplot(aes(frame, i, fill = n_pip)) +
  geom_raster() +
  scale_fill_gradient(low = "white", high = HITS_BLUE, breaks = c(0, 1, 2, 3)) +
  guides(fill = guide_legend(show.limits = TRUE)) +
  labs(
    x = "time [ns]",
    y = "residue index",
    fill = expression(n[pip[2]])
  ) +
  theme(
    legend.position = "bottom", legend.key = element_rect(color = "black")
  ) +
  lims(y = c(0, 400))

anot <- ggplot(ferm_annotation) +
  aes(
    xmin = 0, xmax = 10,
    ymin = imin, ymax = imax,
    fill = domain) +
  geom_rect(show.legend = FALSE) +
  geom_hline(data = ferm_pip_sites, aes(yintercept = ri), color = "red") +
  geom_text(
      aes(x = 5, y = (imin + imax) / 2, label = domain),
      color = "black"
  ) +
  theme_void() +
  scale_fill_discrete(type = ferm_colors)

(plt | anot) +
  plot_layout(widths = c(25, 1)) +
  plot_annotation()

Code 
ferm_interacting |> 
  filter(run != 7) |> 
  ggplot(aes(frame, i, fill = n_pip)) +
  geom_raster() +
  facet_wrap(~ run, labeller = label_both) +
  scale_fill_gradient(low = "white", high = HITS_BLUE, breaks = c(0, 1, 2, 3)) +
  guides(fill = guide_legend(show.limits = TRUE)) +
  labs(
    x = "time [ns]",
    y = "residue index",
    fill = expression(n[pip[2]])
  ) +
  theme(legend.position = "bottom",
        panel.background = element_rect(color = "black", fill = NA))

Code 
plt <- ferm_interacting |> 
  group_by(i) |> 
  summarise(n_pip = mean(n_pip)) |> 
  left_join(sequence) |> 
  left_join(aa, by = c('res' = '1')) |>
  update(1, charge, "Positive") |>
  ggplot(aes(x = i, y = n_pip)) + 
  annotate(geom = "rect",
           xmin = 140, xmax = 164,
           ymin = 0, ymax = Inf,
           color = NA,
           fill = "black",
           alpha = 0.05) +
  geom_col(aes(fill = charge),
           alpha = 1,
           width = 0.8) +
  geom_text(aes(label = res, color = charge),
            fontface = "bold",
            vjust = -0.2,
            data = ~ filter(.x, n_pip > 0.1), show.legend = FALSE) +
  scale_fill_manual(
    values = CHARGE_COLORS, aesthetics = c("fill", "color"),
    name = "Net Charge at pH 7.4"
  ) +
  scale_y_continuous(expand = expansion(mult = c(0, 0.1))) +
  scale_x_continuous(breaks = seq(0, 400, 50)) +
  labs(y = expression(n[PIP[2]])) +
  coord_cartesian(clip = "off", xlim = c(0, 400)) +
  labs(
    x = "residue index",
    y = expression(bar(n)[PIP[2]])
    ) +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    legend.position = c(0, 1),
    legend.justification = c(0, 1),
    legend.title = element_text(size = 12)
  )

grobs <- ggplotGrob(plt)

plt <- plt +
guides(
    color = guide_colorbar(
      barheight = unit(0.7, "grobheight", data = grobs$grobs),
      frame.colour = "black"
    )
)

anot <- ggplot(ferm_annotation) +
  aes(
    ymin = 0, ymax = 10,
    xmin = imin, xmax = imax,
    fill = domain) +
  geom_rect(show.legend = FALSE) +
  geom_vline(data = ferm_pip_sites, aes(xintercept = ri), color = "red") +
  geom_text(aes(y = 5, x = (imin + imax) / 2, label = domain),
            color = "black") +
  theme_void() +
  coord_cartesian(clip = "off", xlim = c(0, 400)) +
  scale_fill_discrete(type = ferm_colors)


(plt / anot) +
  plot_layout(heights = c(20, 1)) +
  plot_annotation()

Code 
f0f1_interacting <- from_first_contact |>
  group_by(i) |> 
  summarise(n_pip = mean(n_pip))

  ferm_interacting |> 
  group_by(i) |> 
  summarise(n_pip = mean(n_pip)) |> 
  left_join(sequence) |> 
  left_join(aa, by = c('res' = '1')) |>
  update(1, charge, "Positive") |>
  left_join(f0f1_interacting, by = c('i'),
      suffix = c('_ferm', '_f0f1')) |>
  replace_na(list(`n_pip.y` = 0)) |>
  filter(between(i, 135, 170)) |>
  pivot_longer(
      starts_with('n_pip'),
      names_to = 'simulation',
      names_prefix = 'n_pip_',
      values_to = 'n_pip') |>
  group_by(simulation) |>
  mutate(n_pip = n_pip / max(n_pip, na.rm = TRUE)) |>
  ungroup() |>
  ggplot(aes(x = i, y = n_pip)) + 
  annotate(geom = "rect",
      xmin = 140, xmax = 164,
      ymin = 0, ymax = Inf,
      color = NA,
      fill = "black",
      alpha = 0.05) +
  geom_col(aes(fill = simulation),
      alpha = 1,
      width = 0.8,
      position = 'dodge2') +
  geom_text(aes(label = res),
      fontface = "bold",
      vjust = -0.2,
      data = ~ filter(.x, n_pip > 0.1) |>
        group_by(i) |> slice_max(n_pip),
      show.legend = FALSE) +
  scale_fill_manual(
      values = c(HITS_BLUE, HITS_GREEN), aesthetics = c("fill"),
      name = "Simulation System"
      ) +
  scale_y_continuous(expand = expansion(mult = c(0, 0.1))) +
  scale_x_continuous(breaks = seq(140, 170, 4)) +
# coord_cartesian(clip = "off", xlim = c(0, 400)) +
  labs(
      x = "residue index",
      y = expression(paste('standardized ', bar(n)[PIP[2]]))
      ) +
  theme(
      axis.text.x = element_text(angle = 45, hjust = 1),
      legend.position = c(0, 1),
      legend.justification = c(0, 1),
      legend.title = element_text(size = 12)
      )

SI

RMSF

Root mean squared fluctiation of the residues of the FERM domain in equilibrium simulaiton compared to the RMSF of the structure ensemble of the NMR structure of F1 (pdb id 2KC2).

Code 
ferm_rmsf |>
  ggplot(aes(residue, rmsd)) +
  annotate(geom = "rect",
    xmin = 135, xmax = 168,
    ymin = 0, ymax = Inf,
    color = NA,
    fill = "black",
    alpha = 0.05) +
  geom_line() +
  geom_line(data = nmr_rmsf, color = HITS_BLUE) +
  geom_rect(
    data = ferm_annotation,
    inherit.aes = FALSE,
    aes(xmin = imin, xmax = imax, fill = domain),
    ymin = 0, ymax = 0.03
  ) +
  labs(
    x = "residue index",
    y = "rmsf [nm]"
  ) +
  scale_fill_discrete(type = ferm_colors) +
  guides(fill = "none")