DIC begins to decline gradually and then decreases sharply at K = 3. It goes back to K = 4 then continues to gradually decline. In summary, there is a peak down to K = 3. Is it a real signal for genetic structure? (See attached picture)
Gabor Borgulya
This looks odd to me. Did your clustering algorithm stop with a result for K=3 comparable to the results for other values of K? If the clustering result for K=3 is qualitatively different from the others (particularly results of K=2 and K=4) you may get the explanation by examining the differences.
2 votes
1 thanks
Bertrand Mallet
Thank you for your answer. There indeed has a problem with K = 3. I made an error in calculating the average DIC on the 10 runs. The curve now has a normal appearance, without breaking.
0 votes
0 thanks
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics
- Genetic Analysis
Similar questions and discussions