Parameter | Definition |
---|---|

HYPERLINKED GO CATEGORY |
(unique) GO category number concatenated to (nonunique) category name - hyperlinked to the AmiGO browser |

HYPERLINKED GENE NAME |
a gene name or gene symbol that is associated with the GO category in the GO CATEGORY column - hyperlinked to Entrez |

TOTAL GENES |
the total number of genes within this category, counting (without duplication) all the genes of all of its descendant categories |

CHANGED GENES |
the number of changed (underexpressed plus overexpressed) genes within this category, counting (without duplication) the changed genes of all of its descendant categories |

UNDEREXPRESSED GENES |
the number of underexpressed genes within this category, counting (without duplication) the underexpressed genes of all of its descendant categories |

OVEREXPRESSED GENES |
the number of overexpressed genes within this category, counting (without duplication) the underexpressed genes of all of its descendant categories |

ENRICHMENT |
informally, the proportion of changed genes n the category relative to the expected proportion: the ratio of changed genes in the category divided by the total number of genes in the category, divided by the same ratio for the entire microarray [depending on the heading in the previous column, changed can refer to any of (changed, underexpressed, overexpressed)] |

LOG10(p) |
the base 10 logarithm of the one-sided Fisher exact p value (uncorrected for multiple comparisons; for full details of the statistical considerations and the null hypothesis, download the PDF version of the original GoMiner article; note that the one-sided rather than the two-sided test is used in this table) |

CUMULATIVE NUMBER OF CATEGORIES |
the cumulative total of the number of categories having a p value less than or equal to that in the previous column |

CUMULATIVE RANDOMS LOWER BOUND |
the mean cumulative number of categories in the random controls (see next column) having a p value less than or equal to that in the LOG10(p) column minus the standard deviation for the random controls |

CUMULATIVE RANDOMS MEAN |
the mean cumulative number of categories in the random controls having a p value less than or equal to that in the LOG10(p) column |

CUMULATIVE RANDOMS UPPER BOUND |
the mean cumulative number of categories in the random controls (see previous column) having a p value less than or equal to that in the LOG10(p) column plus the standard deviation for the random controls |

FDR |
the one-sided Fisher exact p value corrected for multiple comparisons [this is computed in a simple manner by resampling the total genes on the microarray, using these as a proxy for the changed genes in GoMiner, and comparing the distribution of p values in all the categories for the real data and the resampled data; the corrected p value is computed as (RANDOMS MEAN)/(NUMBER OF CATEGORIES) which is an approximation to the fraction of categories that, by random chance, would have had a p value that was as low as that observed for the real data; note that 'changed' can refer to any of (changed, underexpressed, overexpressed)] |

Some files in PDF format and require Adobe Acrobat Reader to be viewed. Click here to download.

GoMiner™ is a development of the Genomics and Pharmacology Facility, Developmental Therapeutics Branch (DTB), Center for Cancer Research (CCR), National Cancer Institute (NCI).

We would like to hear from you. You can reach the team via email.