NAMA : NPM :

R. Nindya Kartika Rachim 140610090075

1. Gene CD33. Use grep to nd the index of the important gene CD33 among the list of characters golub.gnames. For each test below formulate the null hypothesis, the p-value and your conclusion.
(a) Test the normality of the ALL and AML expression values. To perform computations on the expressions of this gene we need to know its row index. This can obtained by the grep function : > grep("CD33",golub.gnames[,2]) [1] 808 From using the function grep, we can find the result, The expression values of antigen CD33 are available at golub[808,] and further information on it by golub.gnames[808,]. The expression values of gene CD33 from the ALL patients can now be printed to the screen, as follows: > golub[808,gol.fac=="ALL"] [1] -0.57277 -1.38539 -0.47039 -0.41469 -0.15402 -1.21719 -1.37386 [8] -0.52956 -1.10366 -0.74396 -0.97673 -0.00787 -0.99141 -1.05662 [15] -1.39503 -0.73418 -0.67921 -0.87388 -0.82569 -1.12953 -0.75991 [22] -0.92231 -1.13505 -1.46474 -0.59614 -1.04821 -1.23051 One of method to visualize data is by dividing the range of data values into a number of intervals is histogram, and can be printed to the screen, as follow : > hist(golub[808, gol.fac=="ALL"])

To test the hypothesis that the ALL gene expression values of CD33 from Golub et al. are normally distributed, the Shapiro-Wilk test can be used as follows. H0 H1 : : : Normal Abnormal 0,05

stat uji : > shapiro.test(golub[808, gol.fac=="ALL"]) Shapiro-Wilk normality test

data: golub[808, gol.fac == "ALL"] W = 0.9696, p-value = 0.592

From the computation, we get p-value = 0.592 > : 0,05. So the hypothesis is accepted, we can say that the data ALL is Normal.

The expression values of gene CD33 from the AML patients can now be printed to the screen, as follows: > golub[808,gol.fac=="AML"] [1] -0.38605 0.50814 0.70283 1.05902 0.38602 -0.19413 1.10560 [8] 0.76630 0.48881 -0.13785 -0.40721 One of method to visualize data is by dividing the range of data values into a number of intervals is histogram, and can be printed to the screen, as follow :

> hist(golub[808, gol.fac=="AML"])

To test the hypothesis that the ALL gene expression values of CD33 from Golub et al. are normally distributed, the Shapiro-Wilk test can be used as follows. H0 H1 : : : Normal Abnormal 0,05

stat uji : > shapiro.test(golub[808, gol.fac=="AML"]) Shapiro-Wilk normality test

data: golub[808, gol.fac == "AML"] W = 0.9121, p-value = 0.2583

From the computation, we p-value = 0.2583 > : 0,05. So the hypothesis is accepted, we can say that the data AML is Normal.

(b) Test for the equality of variances H0 H1 : : : Variance two population are same Variance two population are not same 0,05

stat uji : The null hypothesis for gene CD33 that the variance of the ALL patients equals that of the AML patients can be tested by the built-in-function var.test, as follows. > var.test(golub[808,] ~ gol.fac) F test to compare two variances

data: golub[808, ] by gol.fac F = 0.4605, num df = 26, denom df = 10, p-value = 0.1095 alternative hypothesis: true ratio of variances is not equal to 1 95 percent confidence interval: 0.1376700 1.1923646 sample estimates: ratio of variances 0.4604523

From the computation, we get p-value = 0.1095 > : 0,05. So the hypothesis is accepted. we can say that Variance two population are same

(c) Test for the equality of the means by an appropriate t-test H0 H1 : = 0 : 0 : 0,05

stat uji : we test = 0 against 0 by the built-in-function t-test. Recall that the corresponding gene expression values are collected in row 808 of the golub > t.test(golub[808,] ~ gol.fac, var.equal=TRUE) Two Sample t-test

data: golub[808, ] by gol.fac t = -7.9813, df = 36, p-value = 1.773e-09 alternative hypothesis: true difference in means is not equal to 0 95 percent confidence interval: -1.5487898 -0.9211602

sample estimates: mean in group ALL mean in group AML -0.8812041 0.3537709

From the computation, we get , p-value = 1.773e-09 < : 0,05. So the hypothesis is rejected. we can say that the mean of two population are not same.

(d) Is the experimental effect strong?

From the computation, we get t = -7.9813 , its quite extreme

2. Gene "MYBL2 V-myb avian myeloblastosis viral oncogene homolog- like 2" has its expression values in row 1788.
(a) Use a boxplot to construct a hypothesis about the experimental effect. > boxplot(golub[1788,] ~ gol.fac)

(b) Test for the equality of means by an appropriate t-test. H0 H1 : = 0 : 0 : 0,05

stat uji : we test = 0 against 0 by the built-in-function t-test. Recall that the corresponding gene expression values are collected in row 1788 of the golub > t.test(golub[1788,] ~ gol.fac, var.equal=TRUE)

Two Sample t-test data: golub[1788, ] by gol.fac t = -0.178, df = 36, p-value = 0.8597 alternative hypothesis: true difference in means is not equal to 0 95 percent confidence interval: -0.6874315 0.5764734 sample estimates: mean in group ALL mean in group AML -0.3046481 -0.2491691

From the computation, we get , p-value = 0.8597 > : 0,05. So the hypothesis is accepted. we can say that the mean of two population are same. 3. HOXA9. Gene "HOXA9 Homeo box A9" with expression values in row 1391, can cause leukemia (Golub et al., 1999).
(a) Test the normality of the expression values of the ALL patients. To test the hypothesis that the ALL gene expression values of HOXA9 Homeo box A9" from Golub et al. are normally distributed, the Shapiro-Wilk test can be used as follows. H0 H1 : : : Normal Abnormal 0,05

stat uji : > shapiro.test(golub[1391, gol.fac=="ALL"]) Shapiro-Wilk normality test data: golub[1391, gol.fac == "ALL"] W = 0.5831, p-value = 1.318e-07

From the computation, we p-value = 1.318e-07 < : 0,05. So So the hypothesis is rejected, we can say that the data AML is not Normal.
(b) Test for the equality of means by an appropriate t-test. H0 H1 : = 0 : 0 : 0,05

stat uji :

The null hypothesis that the expression values for gene HOXA9 Homeo box A9" are equally distributed for the ALL patients and the AML patients can be tested by the built-in-function wilcox.test, as follows. > wilcox.test(golub[1391,] ~ gol.fac) Wilcoxon rank sum test data: golub[1391, ] by gol.fac W = 34, p-value = 7.923e-05 alternative hypothesis: true location shift is not equal to 0

From the computation, we get p-value = 7.923e-05 < : 0,05. So the hypothesis is rejected. we can say that the mean of two population are not same.