Welcome to Sample Surfer
Sample Surfer is a program designed to select samples (from a large gene expression data set) that are similar to a given gene signature. Input data are table of gene/protein expression data (samples in columns, genes in rows) and gene signature (often from 5 to 100-200 genes, genes in rows). Three similarity criteria can be use: Pearson Correlation (linear), Spearman Correlation (rank) and Euclidean Distance.
Useful tip: using UCSC Cancer Genomics Browser you can download 720 cancer datasets from TCGA (11,000 samples across over 20 cancer types), CCLE and several major publications. Click “Add Dataset”, select a processed cancer type and dataset that you need and click a button “Download processed dataset” on the left from the heatmap (see more instructions in here). Another source of functional genomics data is ArrayExpress at EBI (you can download both processed and raw data – 58592 experiments total).
If you used Sample Surfer before 8/6/2015, the column labels (the gene names) on the table of Signature Profiles (the second table in your results) may have been inaccurate. If you used the data from that table, you should redo the calculations now that that issue has been fixed. I apologize for this error.
This site was made by Rachel Berman, an intern at the Sanford Burnham Prebys Medical Discovery Institute La Jolla Bioinformatics core with help and guidance from Andrew Hodges and Alexei Eroshkin.
Any questions, suggestions, or issues with it can be emailed to rberman@SBPDiscovery.org