R Markdown

Read in data:

library(readr)
acc_values <- as.data.frame(read_tsv("two_classif.txt"))
## Parsed with column specification:
## cols(
##   dataset = col_character(),
##   algorithm = col_character(),
##   accuracy = col_double()
## )
acc_values4 <- as.data.frame(read_tsv("survival_results.txt"))
## Parsed with column specification:
## cols(
##   dataset = col_character(),
##   algorithm = col_character(),
##   cindex = col_double()
## )
acc_values5 <- as.data.frame(read_tsv("classif_results.txt"))
## Parsed with column specification:
## cols(
##   dataset = col_character(),
##   algorithm = col_character(),
##   accuracy = col_double()
## )

Student’s t-test on classification analysis results for RNA-seq and microarray datasets:

RNAseq_results <- acc_values[acc_values$dataset %in% "RNAseq", 3]
Microarray_results <- acc_values[acc_values$dataset %in% "Microarray", 3]
t.test(RNAseq_results, Microarray_results, var.equal=TRUE, paired=FALSE)
## 
##  Two Sample t-test
## 
## data:  RNAseq_results and Microarray_results
## t = -0.42497, df = 6, p-value = 0.6857
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.06990834  0.04921868
## sample estimates:
## mean of x mean of y 
## 0.8258621 0.8362069

Student’s t-test on survival analysis results for RNA-seq and microarray datsets:

RNAseq_survival_results <- acc_values4[acc_values4$dataset %in% "RNAseq", 3]
Microarray_survival_results <- acc_values4[acc_values4$dataset %in% "Microarray", 3]
t.test(RNAseq_survival_results, Microarray_survival_results, var.equal=TRUE, paired=FALSE)
## 
##  Two Sample t-test
## 
## data:  RNAseq_survival_results and Microarray_survival_results
## t = 0.0056557, df = 6, p-value = 0.9957
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.07362079  0.07396191
## sample estimates:
## mean of x mean of y 
## 0.8246590 0.8244884

One-way ANOVA test comparing the combined dataset with the RNA-seq and microarray datasets in the classificaiton analysis. Displaying the mean adjusted p-value.

anovaFit2 <- aov(accuracy ~ dataset, data=acc_values5[1:12,])
class_3_anova <- TukeyHSD(anovaFit2)
class_3_anova <- as.data.frame(class_3_anova[[1]])
class_3_anova
##                             diff         lwr        upr     p adj
## Microarray-Combined  0.003448276 -0.07279982 0.07969637 0.9912574
## RNAseq-Combined     -0.006896552 -0.08314465 0.06935154 0.9655883
## RNAseq-Microarray   -0.010344828 -0.08659292 0.06590327 0.9245996
mean(class_3_anova[,4])
## [1] 0.9604818

One-way ANOVA test comparing the combined datasets with the RNA-seq and microarray datasets in the survival analysis. Displaying the mean adjusted p-value.

surv_3_anova <- TukeyHSD(aov(cindex ~ dataset , data = acc_values4[1:12,]))
surv_3_anova <- as.data.frame(surv_3_anova[[1]])
surv_3_anova
##                             diff         lwr        upr     p adj
## Microarray-Combined 0.0057939359 -0.07777432 0.08936219 0.9796012
## RNAseq-Combined     0.0059644945 -0.07760376 0.08953275 0.9783989
## RNAseq-Microarray   0.0001705586 -0.08339770 0.08373882 0.9999821
mean(surv_3_anova[,4])
## [1] 0.9859941

One-way ANOVA test comparing classification analysis results of all datasets against each other. Displaying the mean adjusted p-value.

class_all_anova <- TukeyHSD(aov(accuracy ~ dataset, data = acc_values5))
class_all_anova <- as.data.frame(class_all_anova[[1]])
class_all_anova
##                               diff         lwr        upr     p adj
## Fifty-nine-Combined    0.001724138 -0.07165518 0.07510345 0.9998709
## Microarray-Combined    0.003448276 -0.06993104 0.07682759 0.9989745
## RNAseq-Combined       -0.006896552 -0.08027587 0.06648276 0.9920155
## Microarray-Fifty-nine  0.001724138 -0.07165518 0.07510345 0.9998709
## RNAseq-Fifty-nine     -0.008620690 -0.08200000 0.06475863 0.9847166
## RNAseq-Microarray     -0.010344828 -0.08372414 0.06303449 0.9742281
mean(class_all_anova[,4])
## [1] 0.9916128

One-way ANOVA test comparing survival analysis results of all datasets against each other. Displaying the mean adjusted p-value.

surv_all_anova <- TukeyHSD(aov(cindex ~ dataset, data = acc_values4))
surv_all_anova <- as.data.frame(surv_all_anova[[1]])
surv_all_anova
##                                diff         lwr        upr     p adj
## Fifty-nine-Combined   -0.0010723135 -0.09601688 0.09387226 0.9999856
## Microarray-Combined    0.0057939359 -0.08915063 0.10073851 0.9977682
## RNAseq-Combined        0.0059644945 -0.08898008 0.10090906 0.9975674
## Microarray-Fifty-nine  0.0068662494 -0.08807832 0.10181082 0.9963084
## RNAseq-Fifty-nine      0.0070368080 -0.08790776 0.10198138 0.9960307
## RNAseq-Microarray      0.0001705586 -0.09477401 0.09511513 0.9999999
mean(surv_all_anova[,4])
## [1] 0.9979434