Similarity values obtained by G-SESAME
| | PMI40 | SEC53 | GAL1 | GAL7 | GAL10 | UGP1 | YHL012W |
| PMI40 | | 0.338 | 0.192 | 0.173 | 0.338 | 0.173 | 0.283 |
| SEC53 | | | 0.192 | 0.173 | 0.338 | 0.173 | 0.283 |
| GAL1 | | | | 0.311 | 0.195 | 0.311 | 0.500 |
| GAL7 | | | | | 0.173 | 0.746 | 0.254 |
| GAL10 | | | | | | 0.173 | 0.283 |
| UGP1 | | | | | | | 0.254 |
| YHL012W | | | | | | | |
Clustering results obtained by G-SESAME
| Threshold | Initial | 0.740 | 0.500 | 0.330 | 0.310 | 0.280 |
Clustering
Result |
|
|
|
|
|
YHL012W
GAL1
UGP1
GAL7
GAL10
SEC53
PMI40
|
|
Similarity values obtained by Resnik's method
| | PMI40 | SEC53 | GAL1 | GAL7 | GAL10 | UGP1 | YHL012W |
| PMI40 | | 1.378 | 0.919 | 0.281 | 1.378 | 0.281 | 0 |
| SEC53 | | | 0.919 | 0.281 | 1.378 | 0.281 | 0 |
| GAL1 | | | | 2.095 | 1.034 | 2.095 | 0.187 |
| GAL7 | | | | | 0.281 | 4.872 | 0 |
| GAL10 | | | | | | 0.281 | 0 |
| UGP1 | | | | | | | 0 |
| YHL012W | | | | | | | |
Clustering results obtained by Resnik's method
| Threshold | Initial | 4.872 | 2.072 | 1.372 | 0.972 | 0.172 |
Clustering
Result |
|
|
|
|
GAL10
SEC53
PMI40
UGP1
GAL7
GAL1
|
|
GAL10
SEC53
PMI40
UGP1
GAL7
GAL1
YHL012W
|
|
Annotation information for GAL1 and YHL012W:
Red terms annotate gene GAL1. Green term annotates gene YHL012W.
Annotation information for GAL1 and GAL10:
Red terms annotate gene GAL1. Green terms annotate gene GAL10.
Analysis:
The annotation information for GAL1, YHL012W and GAL10 in the above figures show that the minimum distance between the GO terms annotating
GAL1 and YHL012W respectively is 1. On the other hand, the minimum distance between the GO terms annotating GAL1 and GAL10 respectively is 6.
Therefore, the similarity between GAL1 and YHL012W should be greater than the similarity between GAL1 and GAL10 based on the annotation information.
On the other hand, the pathway annotation shows that genes UGP1 and YHL012W are both annotated with the same EC number 2.7.7.9. Therefore,
it is reasonable that these two genes have similarity in their molecular functions.
The similarity values and clustering results obtained by G-SESAME are consistent with these observations.
However, based on Resnik's method the similarity between GAL1 and GAL10 is greater than the similarity between GAL1 and YHL012W,
and the similarity between UGP1 and YHL012W is 0. All these similarity measures are inconsistent
with the annotation information.
Conclusion:
G-SESAME is better.