Here are the links to download datasets used in development of various models of cancerUBM: |
cancerUBM Dataset | ||
|---|---|---|
| Dataset | Description | Download |
fset173 | This dataset contain 173 Weka selected features from 5605 peptides from urinary peptide databaseof 1525 cancer and 1503 healthy samples. This dataset is log transformed by after adding 1 and converted to propensities as described in Materials and Methods. | P N |
fset40 | This dataset contain 40 peptides out of which 20 are more frequent in Cancer samples and 20 are more frequent in Healthy samples.These values are also log-transformed Probability values. | P N |
fset61 | Out of 173 peptides, spectra were selected for which peptide and protein information was available.There were such 61 peptides and model were developed using intensities of these 61 peptides.This data is also log-transformed propensity values. | P N |
fset69 | Out of 5605 peptides, protein information was available for 953 peptides from total 5605 spectra of peptides. We mapped these 953 peptides on human proteins and these peptides map to 118 proteins. There are 69 proteins out of 118 proteins that have more than two peptides fragments. We called this as protein feature set and referred as fset69. This data is log transformed after adding 1. | P_MEAN N_MEAN P_MEDIAN N_MEDIAN P_MAX N_MEDIAN |
fset9 | We also mapped 61 peptides in fset61 on proteins and obtained nine proteins have two or more peptides. We used these nine proteins for developing protein based models and called these proteins as feature set fset9. | P_MEAN N_MEAN P_MEDIAN N_MEDIAN P_MAX N_MAX |
Nine sets for nine Proteins | These are the nine datasets for individual models on nine proteins.The data is log-transformed. | P_ALPHA N_ALPHA P_COL11 N_COL11 P_COL13 N_COL13 P_COL18 N_COL18 P_COL21 N_COL21 P_FIBRO N_FIBRO P_HEM N_HEM P_PRO N_PRO |