## Last updated: Wed Jan 22 19:44:41 2014
NOTE: Always verify figure by manual calculation of a subset in Excel
BQCV All data
1.1 Barplots
1.2 Stats
Nosema All data
2.1 Barplots
2.2 Stats
My data
3.1 Barplots
3.2 Stats
Pan Dudek
4.1 Barplots
4.2 Stats
Pulawy Olszyn Nosema
5.1 Barplots
5.2 Stats
5.3 Summaries
## 'data.frame': 239 obs. of 8 variables:
## $ well : int 1 2 3 4 5 6 7 8 9 10 ...
## $ days : int 7 7 10 12 13 15 16 16 17 20 ...
## $ bqcv : num 15800 218000 12500 46100 158000 15800 30000 50000 71500 3800000 ...
## $ spores : num 0.6 0.71 0.64 1.3 0.97 1.61 1.53 0.5 1.14 2 ...
## $ cage : Factor w/ 3 levels "1","2","3": 1 1 1 1 1 1 1 1 1 1 ...
## $ virus : Factor w/ 2 levels "with BQCV","without BQCV": 1 1 1 1 1 1 1 1 1 1 ...
## $ nosema : Factor w/ 2 levels "N.apis","N.ceranae": 1 1 1 1 1 1 1 1 1 1 ...
## $ log10bqcv: num 4.2 5.34 4.1 4.66 5.2 ...
Barplots for Nosema spores, log10 BQCV copies and Survival days summarised by Nosema species, BQCV presence and cages. The Mean and SD are computed over all individuals in each cage.
Barplot combining Nosema spores, BQCV titres and survival days. Mean and SD computed over all individuals in all cages.
# Stat Stats are usually wilcoxon tests to compare if values in groups are significantly different. A boxplot is shown to visually compare groups. And the test
# results are shown underneath. Blue dot denotes mean. Red + denotes the actual data points.
# 1.2.1 Is there sig diff in nosema spores between colonies with and without BQCV? Yes. the difference is significant. Strange that its sig, since it is not
# visible on the boxplot.
cnt <- ddply(b1, .(virus), summarise, count = fn3(spores))$count
boxplot(b1$spores ~ b1$virus)
stripchart(b1$spores ~ b1$virus, method = "jitter", jitter = 0.3, add = T, col = "red", vert = T, pch = "+")
dp1 <- ddply(b1, .(virus), summarise, spores = fn1(spores))
points(x = 1:2, y = dp1$spores, pch = 16, col = "steelblue", cex = 2)
axis(at = 1:length(cnt), labels = cnt, side = 1, line = 1, cex.axis = 0.8, tick = F)
wilcox.test(spores ~ virus, data = b1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: spores by virus
## W = 5360, p-value = 0.0008672
## alternative hypothesis: true location shift is not equal to 0
# 1.2.2 Is there sig diff in nosema spores between colonies with and without BQCV in both nosema subspecies? In samples with N. ceranae, withoutBQCV produces sig
# higher spores.
cnt <- ddply(b1, .(nosema, virus), summarise, count = fn3(spores))$count
boxplot(b1$spores ~ b1$virus + b1$nosema)
stripchart(b1$spores ~ b1$virus + b1$nosema, method = "jitter", jitter = 0.3, add = T, col = "red", vert = T, pch = "+")
dp1 <- ddply(b1, .(nosema, virus), summarise, spores = fn1(spores))
points(x = 1:4, y = dp1$spores, pch = 16, col = "steelblue", cex = 2)
axis(at = 1:length(cnt), labels = cnt, side = 1, line = 1, cex.axis = 0.8, tick = F)
b1.1 <- subset(b1, as.character(b1$nosema) == "N.apis")
b1.1$nosema <- b1.1$nosema[drop = T]
b1.2 <- subset(b1, as.character(b1$nosema) == "N.ceranae")
b1.2$nosema <- b1.2$nosema[drop = T]
wilcox.test(spores ~ virus, data = b1.1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: spores by virus
## W = 1565, p-value = 0.2183
## alternative hypothesis: true location shift is not equal to 0
wilcox.test(spores ~ virus, data = b1.2)
##
## Wilcoxon rank sum test with continuity correction
##
## data: spores by virus
## W = 1160, p-value = 0.001195
## alternative hypothesis: true location shift is not equal to 0
# 1.2.3 Is there sig diff in BQCV titres between the two nosema species? No
boxplot(b1$log10bqcv ~ b1$nosema)
wilcox.test(log10bqcv ~ nosema, data = b1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: log10bqcv by nosema
## W = 1431, p-value = 0.072
## alternative hypothesis: true location shift is not equal to 0
# 1.2.4 Is there sig diff in survival days between samples with and without BQCV? Yes. Samples withBQCV have sig lower survival days.
boxplot(b1$days ~ b1$virus)
wilcox.test(days ~ virus, data = b1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: days by virus
## W = 5801, p-value = 0.01185
## alternative hypothesis: true location shift is not equal to 0
# 1.2.5 Is there sig diff in survival days between nosema species? Yes. Samples with N.ceranae has sig lower survival days.
boxplot(b1$days ~ b1$nosema)
wilcox.test(days ~ nosema, data = b1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: days by nosema
## W = 9552, p-value = 5.757e-06
## alternative hypothesis: true location shift is not equal to 0
# 1.2.6 Is there sig diff in survival days between samples with and without BQCV in both nosema subspecies? In N. apis infected samples, survival days were sig
# lower withBQCV. In samples withoutBQCV, N.ceranae had sig lower survival days.
cnt <- ddply(b1, .(nosema, virus), summarise, count = fn3(days))$count
boxplot(b1$days ~ b1$virus + b1$nosema)
axis(at = 1:length(cnt), labels = cnt, side = 1, line = 1, cex.axis = 0.8, tick = F)
# make sure 1.2.2 has been run before running below
wilcox.test(days ~ virus, data = b1.1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: days by virus
## W = 1420, p-value = 0.0396
## alternative hypothesis: true location shift is not equal to 0
wilcox.test(days ~ virus, data = b1.2)
##
## Wilcoxon rank sum test with continuity correction
##
## data: days by virus
## W = 1542, p-value = 0.2253
## alternative hypothesis: true location shift is not equal to 0
b1.1 <- subset(b1, as.character(b1$virus) == "with BQCV")
b1.1$virus <- b1.1$virus[drop = T]
b1.2 <- subset(b1, as.character(b1$virus) == "without BQCV")
b1.2$virus <- b1.2$virus[drop = T]
wilcox.test(days ~ nosema, data = b1.1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: days by nosema
## W = 2095, p-value = 0.08307
## alternative hypothesis: true location shift is not equal to 0
wilcox.test(days ~ nosema, data = b1.2)
##
## Wilcoxon rank sum test with continuity correction
##
## data: days by nosema
## W = 2614, p-value = 1.743e-05
## alternative hypothesis: true location shift is not equal to 0
# summary(glm(spores~virus,data=b1)) wilcox.test(spores~virus,data=b1) t.test(spores~virus,data=b1)
# Correlation Nosema vs Days in all samples
ggplot(data = b1, aes(x = days, y = spores)) + geom_point() + facet_wrap(~nosema) + stat_smooth(method = lm) + theme_bw() + ggtitle("Nosema vs days in all samples")
# Correlation test
cor.test(b1$days, b1$spores)
##
## Pearson's product-moment correlation
##
## data: b1$days and b1$spores
## t = 19.51, df = 237, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.7309 0.8293
## sample estimates:
## cor
## 0.785
# Correlation BQCV vs Nosema in samples with BQCV
ggplot(data = b1, aes(x = spores, y = log10bqcv)) + geom_point() + stat_smooth(method = lm) + facet_wrap(~nosema) + theme_bw() + guides(size = FALSE) + ggtitle("BQCV vs spores in samples with BQCV")
# Correlation test
cor.test(b1$log10bqcv, b1$spores)
##
## Pearson's product-moment correlation
##
## data: b1$log10bqcv and b1$spores
## t = 12.71, df = 117, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.6741 0.8281
## sample estimates:
## cor
## 0.7617
# Correlation BQCV vs Days in samples with BQCV
ggplot(data = b1, aes(x = days, y = log10bqcv)) + geom_point() + stat_smooth(method = lm) + facet_wrap(~nosema) + theme_bw() + ggtitle("BQCV vs days in samples with BQCV")
# Correlation test
cor.test(b1$days, b1$log10bqcv)
##
## Pearson's product-moment correlation
##
## data: b1$days and b1$log10bqcv
## t = 9.878, df = 117, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.5626 0.7619
## sample estimates:
## cor
## 0.6743
# Correlation after splitting up samples with and without BQCV Run b1.1 and b1.2 above
# Correlation Nosema vs Days in samples without BQCV
ggplot(data = b1.1, aes(x = days, y = spores)) + geom_point() + stat_smooth(method = lm) + facet_wrap(~nosema) + theme_bw() + ggtitle("Nosema vs days in samples with BQCV")
# Correlation test
cor.test(b1.1$days, b1.1$spores)
##
## Pearson's product-moment correlation
##
## data: b1.1$days and b1.1$spores
## t = 11.94, df = 117, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.6475 0.8127
## sample estimates:
## cor
## 0.7411
# Correlation Nosema vs Days in samples without BQCV
ggplot(data = b1.2, aes(x = days, y = spores)) + geom_point() + stat_smooth(method = lm) + facet_wrap(~nosema) + theme_bw() + ggtitle("Nosema vs days in samples without BQCV")
# Correlation test
cor.test(b1.2$days, b1.2$spores)
##
## Pearson's product-moment correlation
##
## data: b1.2$days and b1.2$spores
## t = 16.58, df = 118, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.7731 0.8832
## sample estimates:
## cor
## 0.8364
## 'data.frame': 238 obs. of 7 variables:
## $ well : int 4 5 6 7 8 9 10 11 12 13 ...
## $ copies : num 1.22e+06 1.38e+07 9.77e+06 1.64e+07 3.47e+07 1.11e+07 1.67e+06 2.04e+08 3.32e+07 2.39e+08 ...
## $ days : int 10 11 12 14 15 15 16 17 21 21 ...
## $ cage : Factor w/ 3 levels "1","2","3": 1 1 1 1 1 1 1 1 1 1 ...
## $ nosema : Factor w/ 2 levels "N.apis","N.ceranae": 1 1 1 1 1 1 1 1 1 1 ...
## $ type : Factor w/ 2 levels "mixed","single": 2 2 2 2 2 2 2 2 2 2 ...
## $ log10copies: num 6.09 7.14 6.99 7.21 7.54 ...
Barplots of Nosema spores and survival days showing Mean and SD.
# 2.2.1 Is there sig diff in nosema copies between single and mixed infections? No
cnt <- ddply(n1, .(type), summarise, count = fn3(log10copies))$count
boxplot(n1$log10copies ~ n1$type)
axis(at = 1:length(cnt), labels = cnt, side = 1, line = 1, cex.axis = 0.8, tick = F)
wilcox.test(log10copies ~ type, data = n1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: log10copies by type
## W = 8040, p-value = 0.07065
## alternative hypothesis: true location shift is not equal to 0
# 2.2.2 Is there sig diff in nosema copies between single and mixed infections in both nosema species? In samples infected with N.apis, mixed infections had sig
# higher N. apis copies. In case of single infection, N. ceranae had sig higher copies than N.apis. In case of mixed infection, N. cerane still had sig higher
# copies than N. apis.
cnt <- ddply(n1, .(nosema, type), summarise, count = fn3(log10copies))$count
boxplot(n1$log10copies ~ n1$type + n1$nosema)
axis(at = 1:length(cnt), labels = cnt, side = 1, line = 1, cex.axis = 0.8, tick = F)
n1.1 <- subset(n1, as.character(n1$nosema) == "N.apis")
n1.1$nosema <- n1.1$nosema[drop = T]
n1.2 <- subset(n1, as.character(n1$nosema) == "N.ceranae")
n1.2$nosema <- n1.2$nosema[drop = T]
wilcox.test(log10copies ~ type, data = n1.1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: log10copies by type
## W = 3009, p-value = 4.624e-11
## alternative hypothesis: true location shift is not equal to 0
wilcox.test(log10copies ~ type, data = n1.2)
##
## Wilcoxon rank sum test with continuity correction
##
## data: log10copies by type
## W = 1287, p-value = 0.01033
## alternative hypothesis: true location shift is not equal to 0
n1.1 <- subset(n1, as.character(n1$type) == "single")
n1.1$type <- n1.1$type[drop = T]
n1.2 <- subset(n1, as.character(n1$type) == "mixed")
n1.2$type <- n1.2$type[drop = T]
wilcox.test(log10copies ~ nosema, data = n1.1)
##
## Wilcoxon rank sum test with continuity correction
##
## data: log10copies by nosema
## W = 270.5, p-value = 1.014e-15
## alternative hypothesis: true location shift is not equal to 0
wilcox.test(log10copies ~ nosema, data = n1.2)
##
## Wilcoxon rank sum test with continuity correction
##
## data: log10copies by nosema
## W = 966.5, p-value = 3.138e-05
## alternative hypothesis: true location shift is not equal to 0
## 'data.frame': 42 obs. of 7 variables:
## $ colony : Factor w/ 3 levels "Colony 1","Colony 3",..: 1 2 3 1 2 3 1 2 3 1 ...
## $ spores : num 3.6 33.8 17.6 0 15.2 12.3 7.8 4 5.7 1.4 ...
## $ perinf : int 60 100 26 3 33 20 87 60 47 17 ...
## $ nosema : Factor w/ 2 levels "Mixed","N.ceranae": 2 2 2 2 2 2 2 2 2 2 ...
## $ nosema1: Factor w/ 3 levels "Mixed","N.ceranae",..: 2 2 2 2 2 2 2 2 2 2 ...
## $ season : Factor w/ 3 levels "spring","autumn",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ year : Factor w/ 6 levels "2007","2008",..: 3 3 3 4 4 4 5 5 5 6 ...
What Stats to do with this dataset?
## 'data.frame': 30 obs. of 6 variables:
## $ colony: Factor w/ 6 levels "3","5","6","7",..: 1 2 3 4 5 6 1 2 3 4 ...
## $ spores: num 8.1 0.5 3.1 10.7 4.9 2.3 0.1 NA 0.4 0.8 ...
## $ year : Factor w/ 2 levels "2010","2011": 1 1 1 1 1 1 1 1 1 1 ...
## $ season: Factor w/ 3 levels "spring","autumn",..: 1 1 1 1 1 1 2 2 2 2 ...
## $ nosema: Factor w/ 2 levels "Mixed","N.ceranae": 1 1 1 1 1 1 2 NA 1 1 ...
## $ perinf: num 83 56.6 70 96.6 93.3 56.6 35 NA 13.3 16.6 ...
What Stats to do with this dataset?
## 'data.frame': 840 obs. of 12 variables:
## $ colony : Factor w/ 105 levels "100","101","103",..: 1 2 3 4 5 6 7 8 9 10 ...
## $ spores : num 0.3 2.25 3.25 0 0 5.2 5.6 0 0.5 0 ...
## $ perinf : num 3 15 35 0 0 20 5 0 9 0 ...
## $ nosema : Factor w/ 3 levels "Mixed","N.apis",..: 1 2 1 NA NA 3 1 NA 1 NA ...
## $ nosema1 : Factor w/ 5 levels "Mixed","N.apis",..: 1 2 1 NA NA 3 1 NA 1 NA ...
## $ season : Factor w/ 3 levels "Summer","Autumn",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ year : Factor w/ 3 levels "2009","2010",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ seasyear: Factor w/ 8 levels "Summer 2009",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ status : Factor w/ 2 levels "Alive","Dead": 1 1 1 1 1 1 1 1 1 1 ...
## $ status1 : Factor w/ 8 levels "Dead in winter 2009",..: 8 3 7 4 2 2 2 2 2 2 ...
## $ status2 : Factor w/ 4 levels "2009","2010",..: 4 2 3 3 2 2 2 2 2 2 ...
## $ status3 : Factor w/ 2 levels "Alive","Dead": 1 2 2 2 2 2 2 2 2 2 ...
Note: Y-Axis not fixed
General barplots
Barplots for survival status
What stats to do?
## Season Nosema Counts
## 1 Summer 2009 Mixed 56
## 2 Summer 2009 N.apis 3
## 3 Summer 2009 N.ceranae 3
## 4 Summer 2010 Mixed 14
## 5 Summer 2010 N.ceranae 22
## 6 Summer 2011 Mixed 3
## 7 Summer 2011 N.ceranae 20
