These are a few of my favorite graphs

Introduction

There are always some big papers which one cites left and right, because they’ve generated a very important dataset, or have a very succinct paper. Here, I will be using R to improve upon these original graphs, and also provide an interactive view of them.

Libraries

library(tidyverse)
options(readr.num_columns = 0)
library(ggrepel)
library(ggpubr)
library(plotly)

SEER Cancer Statistics: Cancer Incidence as a Function of Age

seer <- read_lines(
  "data/SEER18_IncidenceByAge_Sex_Race.csv",
  skip=3
  ) %>% 
  head(n=-13) %>% 
  paste(collapse = "\n") %>%
  read_csv(
    na = "-",
    col_names = T,
    col_types = list(
      "Sex"=col_factor(c("Both Sexes", "Female", "Male")),
      "Race/Ethnicity"=col_factor(
        c('All Races (includes Hispanic)', 'American Indian / Alaska Native (includes Hispanic)', 'Asian / Pacific Islander (includes Hispanic)', 'Black (includes Hispanic)', 'Hispanic (any race)', 'Non-Hispanic White', 'White (includes Hispanic)')
        ),
      "Age" = col_factor(
        c('<1', '1-4', '5-9', '10-14', '15-19', '20-24', '25-29', '30-34', '35-39', '40-44', '45-49', '50-54', '55-59', '60-64', '65-69', '70-74', '75-79', '80-84', '85+'), 
        ordered = T
        )
      )
    ) %>% 
  mutate(`Race/Ethnicity` = str_remove(`Race/Ethnicity`, "\\(includes Hispanic\\)"))

seer.plot <- seer %>%
  filter(
    Sex=="Both Sexes",
    !Age %in% c("<1","5-9") # No data
    ) %>% 
  ggplot(aes(x=as.factor(`Age`), y=`Rate per 100,000`, color=`Race/Ethnicity`)) + 
  geom_point() +
  geom_ribbon(
    aes(
      ymin=`Rate Lower Confidence Interval`, 
      ymax=`Rate Upper Confidence Interval`, 
      ),
    alpha=0.2
  ) + 
  scale_y_log10() + 
  labs(
    y="Incidence Rate per 100,000 (log)",
    x="Age at Diagnosis",
    title="SEER18 Cancer Statistics Review, 1975-2015, NCI"
    ) + 
  theme_pubclean() + 
  theme(
    axis.text.x = element_text(angle = 45, vjust = .5),
    legend.position = "bottom"
    )

seer.plot %>% ggplotly()

million.women <- read_csv(
  "data/Height_Cancer_Million_Women_Study_2011.csv",
  col_types = list(
    "Height" = col_factor(
      c('<155', '155', '160', '165', '170', '>=175'), 
      ordered = T
    )
  )
  ) 

million.women.plot <- million.women %>% 
  ggplot(
    aes(x=Height, y=RR, text=str_glue("Sample Size: {N.Women}"))
  ) +
  geom_line(
    aes(group=1)
  )+
  geom_point() +
  geom_ribbon(
    aes(ymin=LCI, ymax=UCI, group=1),
    alpha=0.2
  ) + 
  labs(
    y="Relative Risk of Cancer (vs <155cm)",
    x="Height (cm)",
    title="J Green et al, 2011, Million Women Cancer Study"
    ) + 
  theme_pubclean() + 
  theme(axis.text.x = element_text(angle = 45, vjust = .5))

million.women.plot %>% ggplotly()

## Warning: `group_by_()` is deprecated as of dplyr 0.7.0.
## Please use `group_by()` instead.
## See vignette('programming') for more help
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_warnings()` to see where this warning was generated.

Abegglen2015 <- read_csv("data/Abegglenetal2015CancerTable.csv") %>% 
  mutate(
    logMassLifeSpan=log10(`Adultmass(Kg)`)+log10(`Maximumlifespan(yrs)`)
    ) 

Abegglen2015.plot <- Abegglen2015 %>% 
  ggplot(
    aes(
      x=logMassLifeSpan, 
      y=`%Tumors`,
      weight=`#Necropsies`#,
      # text=str_glue("{CommonName}")
      )
    ) +
  geom_point(aes(text=str_glue("{CommonName}")))+
  geom_errorbar(aes(ymin=`Lowerlimit95%CI`, ymax=`Upperlimit95%CI`), width=.1, alpha=0.2) +
  labs(
    y="% Necropsies with Tumors",
    x="log(Mass(g)xLifespan(yrs))",
    title="Abegglen et al, 2015"
      ) +
  geom_smooth(
    method="lm",
    lty="dashed",
    color="black"
    ) + 
  theme_pubclean()  + 
  ylim(c(-1,80))

## Warning: Ignoring unknown aesthetics: text

Abegglen2015.plot %>% ggplotly()

## `geom_smooth()` using formula 'y ~ x'