Hi,
I am very new to bioinformatics. I have applied 2-3 data mining/ machine learning techniques to detect gene clusters using RNASeq gene expression data. Similarly I also know it is possible to do promoter and pathway analysis or gene ontology on that data using machine learning/data mining algorithms.
1) Can you suggest some very new clustering and/or classification algorithm with very high performance considering replicates at each time point.
2) How can I measure performance of the clustering mechanism that I am using ?
3) What visualisation method would be best for showing the result of clustering?
4) Can you suggest different types of meaningful biological analysis that can be performed on the gene clusters to have a meaningful conclusion on the significance of gene clusters.
It sounds like you're more interested in data science approaches -- I would recommend this book which answers many of the questions you pose: https://jakevdp.github.io/PythonDataScienceHandbook/ -- you can also take the machine learning course on Data Camp which addresses many of the points you raised.
WGCNA is a wonderful tool for discovering gene 'modules' in your data set. Also, I tend to gravitate towards clusterProfiler and enrichr for doing gene set over-representation testing -- this lets you determine what the biological significance of each cluster is.
Thanks a lot sir. Can you propose some RNASeq time series datasets on human conditions ? It would be of great help.