HOME DESIGN MICROARRAY SAMPLES NORMALIZATION  SIGNIFICANCE TESTING SAM  H. CLUSTERING  PAM CLUSTERING  PAM CLASSIFICATION  CONCLUSIONS 




This web page was produced as an assignment for a course on Statistical Analysis of Microarray Data at Pomona College.

Initial Significance Testing:
Data from the two "normal" tissue sample types (skeletal (NORM1) and smooth (NORM2) muscle) were compared to data from the synovial sarcoma samples (SS).

NORM1-SS Comparison:
Below are the spots that correlate to the top ten significant miRNAs from the NORM1-SS comparison, as well as their associated p-values (H0: miRNA expression does not differ between the groups). The miRNAs are not further annotated since they are not genes.

miRNA spot number

miRNA name

p-value

reject H0 at 0.05 sig. level?

reject H0 with Bonferroni-adjusted alpha = 3.2e-05?

377

hsa_miR_133b

1.944130e-10

Yes

Yes

1419

hsa_miR_133a

4.528836e-10

Yes

Yes

1395

hsa_miR_133a

6.363466e-10

Yes

Yes

1499

hsa_miR_1

9.546298e-10

Yes

Yes

1448

hsa_miR_422a

1.230338e-09

Yes

Yes

1137

hsa_miR_206

1.278166e-09

Yes

Yes

695

(blank)

4.977765e-08

Yes

Yes

1472

hsa_miR_422a

5.279171e-08

Yes

Yes

719

(blank)

7.192580e-08

Yes

Yes

197

hsa_miR_214

1.106340e-07

Yes

Yes


For each of the above miRNAs, the null hypothesis is rejected in both adjusted and non-adjusted tests. Thus, we can conclude that these miRNAs are differentially expressed between the Normal Skeletal Muscle samples and the Synovial Sarcoma tumors.

The Volcano Plot below gives a more visual representation of the significance testing for miRNAs. The top ten miRNAs from the table are marked on the graph. Because it plots the log of the fold change versus the log odds, it also provides more information.

What we are especially interested in are miRNAs that are highly differentially expressed (high absolute value of Log Fold Change), but are also statistically significant (High value for Log Odds). Thus, spots that appear in the upper right and left sides of the "volcano" are most interesting to us because they are the spots that are both biologically and statistically significant. The fact that the top ten miRNAs from our table are also in the best position on the graph is promising!


NORM2-SS Comparison:
Below are the spots that correlate to the top ten significant miRNAs from the NORM2-SS comparison, as well as their associated p-values (H0: miRNA expression does not differ between the groups).

miRNA spot number

miRNA name

p-value

reject H0 at 0.05 sig. level?

reject H0 with Bonferroni-adjusted alpha = 3.2e-05?

993

hsa_miR_145

2.433386e-11

Yes

Yes

969

hsa_miR_145

3.215736e-11

Yes

Yes

607

hsa_miR_143

1.009558e-10

Yes

Yes

499

ambi_miR_7510

1.153537e-10

Yes

Yes

583

hsa_miR_143

1.287692e-10

Yes

Yes

523

ambi_miR_7510

3.552259e-10

Yes

Yes

75

hsa_miR_29b

9.751187e-10

Yes

Yes

51

hsa_miR_29b

4.545147e-09

Yes

Yes

1107

hsa_miR_194

5.633542e-08

Yes

Yes

265

hsa_miR_222

9.843991e-08

Yes

Yes


For each of the above miRNAs, the null hypothesis is rejected. Thus, we can conclude that these miRNAs are differentially expressed between the Normal Smooth Muscle samples and the Synovial Sarcoma tumors. One interesting thing to note is that the top ten significant miRNAs differ between the two comparisons, even though each comparison is the tumor type against a "control". This is an important observation with regards to the different types of muscle control.

The Volcano Plot below gives a more visual representation of the significance testing for miRNAs. The top ten miRNAs from the table are marked on the graph. Because it plots the log of the fold change versus the log odds, it also provides more information.

What we are especially interested in are miRNAs that are highly differentially expressed (high absolute value of Log Fold Change), but are also significant (High value for Log Odds). Thus, spots that appear in the upper right and left sides of the "volcano" are most interesting to us because they are the spots that are both biologically and statistically significant. The fact that the top ten miRNAs from our table are also in the best position on the graph is promising!