Clustering large data

Question

[rishi]

I have multiple datasets of genomic data, of several thousand rows and 20 columns. Each dataset is a cell line/tissue type, and each row is a genomic feature.

I wrote my own algorithm which worked on the raw dataset to cluster all the genomic feature rows of the large dataset matrices in a supervised way, to obtain 7 classes. Now I want to validate this with a supervised clustering algorithm, if I can get the somewhat similar 7 classes as well, with the same approach/idea as I used in my clustering.

What are the best ways to do this, and how can I do so?

Answer 1

[Raony Guimarães]

You could use scikit-learn http://scikit-learn.org

http://scikit-learn.org/stable/modules/clustering.html#clustering

Answer 2

[arta]

You can use ConsensusClustering or Gap Statistics in R to compare your classification and unsupervised clustering methods.