scores: Data: Probe set quality scores
In jetset: One-to-One Gene-Probeset Mapping for Affymetrix Human Microarrays

Description Usage Format Details Note Source References See Also Examples

This data set provides gene target and quality scores for each probe set on the corresponding Affymetrix gene expression microarrays.

scores.hgu95av2
scores.hgu133a
scores.hgu133plus2
scores.u133x3p

A data frame with each row corresponding to a probe set, with 4 columns:

EntrezID: Entrez GeneID of the targeted gene (character).
process: Processivity requirement (integer).
specificity: Specificity score (numeric).
coverage: Coverage score (numeric).

If there is a relative majority (plurality) of the probes in a probe set that are specific for a single gene, this is defined as the targeted gene. If no such majority exists, the targeted gene is defined as NA, as are the following scores.

The processivity requirement is the number of consecutive bases that must be synthesized to generate a target that can be detected by the probe set.

The specificity score is the fraction of the probes in a probe set that are likely to detect the targeted gene and unlikely to detect other genes.

The coverage score is the fraction of the splice isoforms belonging to the targeted gene that are detected by the probe set.

The following two scores are not contained in this data, but are calculated from the above scores; to see them use jscores.

The robustness score quantifies robustness against transcript degradation. The robustness score uses the processivity requirement to estimate the signal intensity of a probe set, relative to the ideal case of perfect processivity.

The overall score is the product of the specificity score, coverage score, and robustness score.

All scores can range from 0 to 1. A higher score indicates better (predicted) performance.

Details about the jetset algorithm are available in the vignette.

This data is also available in CSV format from http://www.cbs.dtu.dk/biotools/jetset/

Scores are calculated from BLASTN alignments between probe sequences and Refseq transcript sequences, as described in the vignette and in the reference below.

The Refseq human RNA was downloaded from NCBI on 2017-04-04. The lookups were based on org.Hs.eg.db version 3.4.0.

Qiyuan Li, Nicolai J. Birkbak, Balazs Gyorffy, Zoltan Szallasi and Aron C. Eklund. (2011) Jetset: selecting the optimal microarray probe set to represent a gene. BMC Bioinformatics. 12:474.

jscores for a more convenient way to access this data

  ## Here is the EntrezID for the ESR1 gene
  id <- "2099"
  
  ## Extract the scores for all probe sets detecting ESR1
  scores.hgu95av2[which(scores.hgu95av2$EntrezID == id), ]

  ## Compare to the recommended function 'jscores'
  jscores("hgu95av2", eg = "2099")