title: “Evaluación del módulo Bioestadistica” subtitle: “Procesamiento y Simulación de Datos Biológicos” author: “JESUS AMADO MORINIGO SOSA” date: “Curso 2025-2026” urlcolor: blue output: html_document: toc: no —

0. Lectura de los datos

datos_clinicos <- read.csv("data/datos_clinicos.csv", stringsAsFactors = FALSE)
farmacos       <- read.csv("data/farmacos.csv", stringsAsFactors = FALSE)
metadatos      <- read.csv("data/metadatos.csv", stringsAsFactors = FALSE)

kable(data.frame(
  fichero = c("datos_clinicos","farmacos","metadatos"),
  filas  = c(nrow(datos_clinicos), nrow(farmacos), nrow(metadatos)),
  columnas = c(ncol(datos_clinicos), ncol(farmacos), ncol(metadatos))
))

fichero	filas	columnas
datos_clinicos	1888	4
farmacos	249	2
metadatos	249	4

1. Análisis descriptivo

df <- datos_clinicos %>%
  left_join(farmacos, by = "ID") %>%
  left_join(metadatos, by = "ID")

meta_unicos <- metadatos %>% distinct(ID, .keep_all = TRUE)

kable(data.frame(
  Edad_media = mean(meta_unicos$Age, na.rm = TRUE),
  Peso_medio = mean(meta_unicos$Weight, na.rm = TRUE)
), digits = 2)

Edad_media	Peso_medio
12.73	26.12

kable(summary(df %>% select(Dias, Volumen, Metastasis)))

Dias	Volumen	Metastasis
Min. : 0.0	Min. :22.05	Min. :0.000
1st Qu.: 5.0	1st Qu.:45.00	1st Qu.:0.000
Median :20.0	Median :48.95	Median :1.000
Mean :19.6	Mean :50.45	Mean :1.023
3rd Qu.:30.0	3rd Qu.:56.32	3rd Qu.:2.000
Max. :45.0	Max. :78.57	Max. :4.000

Ratones por sexo

kable(df %>% distinct(ID, Sex) %>% count(Sex))

Sex	n
Female	124
Male	125

Ratones por fármaco

kable(df %>% distinct(ID, Farmaco) %>% count(Farmaco) %>% arrange(desc(n)))

Farmaco	n
Capomulin	25
Ceftamin	25
Infubinol	25
Ketapril	25
Naftisol	25
Placebo	25
Propriva	25
Ramicane	25
Zoniferol	25
Stelasyn	24

Ratones macho por fármaco

kable(df %>% distinct(ID, Farmaco, Sex) %>%
        filter(Sex %in% c("Male","M","macho","Macho")) %>%
        count(Farmaco))

Farmaco	n
Capomulin	12
Ceftamin	12
Infubinol	13
Ketapril	16
Naftisol	12
Placebo	12
Propriva	13
Ramicane	16
Stelasyn	9
Zoniferol	10

Volumen final por fármaco

ultimo <- df %>%
  group_by(ID) %>%
  filter(Dias == max(Dias, na.rm = TRUE)) %>%
  ungroup() %>%
  distinct(ID, .keep_all = TRUE)

resumen_vol <- ultimo %>%
  group_by(Farmaco) %>%
  summarise(
    n = n(),
    min = min(Volumen),
    mediana = median(Volumen),
    media = mean(Volumen),
    max = max(Volumen),
    sd = sd(Volumen),
    var = var(Volumen),
    .groups="drop"
  ) %>%
  arrange(media)

kable(resumen_vol, digits = 2)

Farmaco	n	min	mediana	media	max	sd	var
Ramicane	25	22.05	36.56	36.19	45.22	5.67	32.17
Capomulin	25	23.34	38.13	36.67	47.69	5.72	32.66
Propriva	25	45.00	55.84	56.74	72.46	8.33	69.35
Ceftamin	25	45.00	59.85	57.75	68.92	8.37	69.98
Infubinol	25	36.32	60.17	58.18	72.23	8.60	74.01
Zoniferol	25	45.00	61.84	59.18	73.32	8.77	76.86
Placebo	25	45.00	62.03	60.51	73.21	8.87	78.76
Stelasyn	24	45.00	62.19	61.00	75.12	9.50	90.33
Naftisol	25	45.00	63.28	61.21	76.67	10.30	106.03
Ketapril	25	45.00	64.49	62.81	78.57	9.95	98.92

ggplot(ultimo, aes(x = reorder(Farmaco, Volumen, FUN = median), y = Volumen)) +
  geom_boxplot() +
  coord_flip()

Variación del volumen (desde 45)

ultimo <- ultimo %>% mutate(Variacion = Volumen - 45)

kable(ultimo %>%
        group_by(Farmaco) %>%
        summarise(media_variacion = mean(Variacion),
                  sd_variacion = sd(Variacion),
                  .groups="drop") %>%
        arrange(media_variacion),
      digits = 2)

Farmaco	media_variacion	sd_variacion
Ramicane	-8.81	5.67
Capomulin	-8.33	5.72
Propriva	11.74	8.33
Ceftamin	12.75	8.37
Infubinol	13.18	8.60
Zoniferol	14.18	8.77
Placebo	15.51	8.87
Stelasyn	16.00	9.50
Naftisol	16.21	10.30
Ketapril	17.81	9.95

Metástasis final por fármaco

kable(ultimo %>%
        group_by(Farmaco) %>%
        summarise(media = mean(Metastasis),
                  mediana = median(Metastasis),
                  sd = sd(Metastasis),
                  .groups="drop") %>%
        arrange(media),
      digits = 2)

Farmaco	media	mediana	sd
Ramicane	1.20	1	0.87
Capomulin	1.28	1	0.98
Stelasyn	1.46	1	1.06
Propriva	1.56	1	1.36
Infubinol	1.60	1	1.22
Ceftamin	1.72	1	1.43
Ketapril	1.92	2	1.66
Naftisol	2.00	2	1.35
Zoniferol	2.00	2	1.50
Placebo	2.08	2	1.50

ggplot(ultimo, aes(x = reorder(Farmaco, Metastasis, FUN = median), y = Metastasis)) +
  geom_boxplot() +
  coord_flip()

2. Subconjunto de 4 fármacos

subfarm <- c("Capomulin","Infubinol","Ketapril","Placebo")

ultimo4 <- ultimo %>% filter(Farmaco %in% subfarm)

kable(ultimo4 %>% count(Farmaco))

Farmaco	n
Capomulin	25
Infubinol	25
Ketapril	25
Placebo	25

kable(ultimo4 %>%
        group_by(Farmaco) %>%
        summarise(media_vol = mean(Volumen),
                  sd_vol = sd(Volumen),
                  media_met = mean(Metastasis),
                  sd_met = sd(Metastasis),
                  .groups="drop"),
      digits = 2)

Farmaco	media_vol	sd_vol	media_met	sd_met
Capomulin	36.67	5.72	1.28	0.98
Infubinol	58.18	8.60	1.60	1.22
Ketapril	62.81	9.95	1.92	1.66
Placebo	60.51	8.87	2.08	1.50

p1 <- ggplot(ultimo4, aes(Farmaco, Volumen)) + geom_boxplot()
p2 <- ggplot(ultimo4, aes(Farmaco, Metastasis)) + geom_boxplot()

grid.arrange(p1, p2, ncol=2)

3. ANOVA

modelo_aov <- aov(Volumen ~ Farmaco, data = ultimo4)
kable(as.data.frame(anova(modelo_aov)), digits = 4)

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
Farmaco	3	10915.314	3638.4380	51.1816	0
Residuals	96	6824.529	71.0888	NA	NA

kable(as.data.frame(TukeyHSD(modelo_aov)$Farmaco), digits = 6)

	diff	lwr	upr	p adj
Infubinol-Capomulin	21.510678	15.275454	27.745903	0.000000
Ketapril-Capomulin	26.138624	19.903399	32.373848	0.000000
Placebo-Capomulin	23.840847	17.605622	30.076071	0.000000
Ketapril-Infubinol	4.627946	-1.607279	10.863170	0.218170
Placebo-Infubinol	2.330168	-3.905056	8.565393	0.762787
Placebo-Ketapril	-2.297777	-8.533002	3.937447	0.770412

4. Correlación y Regresión

medias4 <- ultimo4 %>%
  group_by(Farmaco) %>%
  summarise(media = mean(Volumen), .groups="drop") %>%
  arrange(media)

mejor <- medias4$Farmaco[1]

df_best <- ultimo4 %>% filter(Farmaco == mejor)

cor_res <- cor.test(df_best$Weight, df_best$Volumen)

kable(data.frame(
  Farmaco = mejor,
  r = unname(cor_res$estimate),
  p_value = cor_res$p.value
), digits = 6)

Farmaco	r	p_value
Capomulin	0.876706	0

modelo_lm <- lm(Volumen ~ Weight, data = df_best)
kable(as.data.frame(summary(modelo_lm)$coefficients), digits = 6)

	Estimate	Std. Error	t value	Pr(>\|t\|)
(Intercept)	1.448147	4.068291	0.355960	0.725115
Weight	1.750468	0.200265	8.740754	0.000000

ggplot(df_best, aes(Weight, Volumen)) +
  geom_point() +
  geom_smooth(method="lm", se=TRUE)