Clinically Annotated Data for the Ovarian Cancer Transcriptome
curatedOvarianData-package.RdThe curatedOvarianData package provides manually curated clinical data, uniformly processed expression data, and convenience functions for gene expression analysis in patients with ovarian cancer.
Details
| Package: | curatedOvarianData |
| Type: | Package |
| Version: | 1.46.2 |
| Date: | 2023-10-31 |
| License: | Artistic-2.0 |
| Depends: | R (>= 2.10.0), affy |
Please see http://bcb.dfci.harvard.edu/ovariancancer for alterative versions of this package, differing in how redundant probe sets are dealt with. In the curatedOvarianData version, each gene is represented by the gene with maximum mean. In NormalizerVcuratedOvarianData, each gene is represented by the mean of the probesets after removing "noisy" probesets (see the Normalizer function of the Sleipnir library for computational biology), and in FULLVcuratedOvarianData, no collapsing of probe sets is done, but a map is provided to allow the user to do so by their method of choice through featureData(eset).
In the "Available sample meta-data" sections of each dataset, please refer to the following key.
For "sample_type": tumor / metastatic / adjacentnormal / healthy / cellline: "healthy" should be only from individuals without cancer, "adjacentnormal" from individuals with cancer, "metastatic" for non-primary tumors.
For "histological_type": ser=serous / endo=endometrioid / clearcell / mucinous, undifferentiated / other / mix. Other includes sarcomatoid, adenocarcinoma, dysgerminoma.
For "primarysite" and for "arrayedsite": ov|ft|other. ov=ovary;ft=fallopian tube
For "summarygrade": low = 1, 2, LMP. High= 3,4,23.
For "summarystage": early = 1,2, 12. late=3,4,23,34.
For "tumorstage": FIGO Stage (I-IV, but coded here as 1-4 to ensure correct ordering in factors). If multiple stages given (eg 34), use the highest.
For "substage": substage (abcd). For cases like ab, bc, use highest given.
For "grade": Grade (1-3): If multiple given, ie 12, 23, use highest given. Most ovarian cancer studies use FIGO grading, with a couple exceptions in this package (Yoshihara and Tothill).
For "pltx": (y/n): patient treated with platin.
For "tax": (y/n): patient treated with taxol.
For "neo": (y/n): patient treated with neoadjuvant treatment.
For "primary_therapy_outcome_success": completeresponse|partialresponse|progressivedisease|stabledisease: response to any kind of therapy (including radiation only).
For "days_to_tumor_recurrence": time to recurrence or last follow-up in days
For "recurrence_status": recurrence censoring variable (recurrence / norecurrence)
For "days_to_death": time to death or last follow-up in days
For "vital_status": Overall survival censoring variable (living / deceased)
For "os_binary": dichotomized overall survival variable as defined by study authors (short / long).
For "relapse_binary": dichotomized relapse variable as defined by study authors (short / long)
For "site_of_tumor_first_recurrence": (metastasis / locoregional / none / locoregional_plus_metastatic). none for no recurrence, na for unknown
For "primary_therapy_outcome_success": (completeresponse / partialresponse / progressivedisease / stabledisease) Response to any kind of therapy (including radiation only).
For "debulking": amount of residual disease (optimal = <1cm, suboptimal=>1cm).
For "percent_normal_cells": Estimated percentage of normal cells. An integer 0-100, or can be >70, <70, etc.
For "percent_stromal_cells": Estimated percentage of stromal cells. An integer 0-100, or can be >70, <70, etc.
For "percent_tumor_cells": Estimated percentage of tumor cells. An integer 0-100, or can be >70, <70, etc.
For "batch": batch variable when available. Hybridization date when Affymetrix CEL files are available.
For "uncurated_author_metadata": Original uncurated data, with each field separated by ///.
Author
Benjamin F. Ganzfried, Steve Skates, Markus Riester, Victoria Wang, Thomas Risch, Benjamin Haibe-Kains, Curtis Huttenhower, Svitlana Tyekucheva, Jie Ding, Ina Jazic, Michael Birrer, Giovanni Parmigiani, Levi Waldron
Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Harvard School of Public Health
Maintainer: Levi Waldron <levi@jimmy.harvard.edu>
Examples
##List all datasets:
data(package="curatedOvarianData")
##
##See the actual template used for syntax checking of clinical metadata:
template.file <- system.file("extdata/template_ov.csv", package = "curatedOvarianData")
template <- read.csv(template.file, as.is=TRUE)
head(template)
#> col.name var.class uniqueness requiredness
#> 1 sample_name character unique required
#> 2 alt_sample_name character unique optional
#> 3 unique_patient_ID character non-unique optional
#> 4 sample_type character non-unique required
#> 5 histological_type character non-unique optional
#> 6 primarysite character non-unique optional
#> allowedvalues
#> 1 *
#> 2 *
#> 3 *
#> 4 tumor|metastatic|borderline|benign|adjacentnormal|healthy|cellline
#> 5 ser|endo|clearcell|mucinous|undifferentiated|other|mix
#> 6 ov|ft|other
#> description
#> 1 primary sample identifier
#> 2 if another identifier is used, for example in supplemental tables
#> 3 Use this column if there are technical replicates. If this column contains non-unique entries, expression values of those arrays will eventually be averaged.
#> 4 healthy should be only from individuals without cancer, adjacentnormal from individuals with cancer, metastatic for non-primary tumors, borderline includes both borderline and LMP tumors, benign for benign tumors.
#> 5 ser=serous;endo=endometrioid;clearcell=mixture of ser+endo. Other includes sarcomatoid, endometroid, papillary serous, unspecified adenocarcinoma, dysgerminoma
#> 6 ov=ovary;ft=fallopian tube