Main
Curriculum Vitae
Publications
Projects

BEN RUBINSTEIN

 SOFTWARE

Below are links to software tools I have had some hand in developing. All of these tools have been made available to the public subject to various particular conditions of use, and can be arranged into the following categories (e.g. see my projects page):

Low-level Microarray Data Analysis: Detecting the Presence of mRNA Transcripts

The software tools in this category are described in the following:

Projects Papers
  • Rubinstein & Speed, Detecting Gene Expression with Oligonucleotide Microarrays, paper in preparation 2005.

  • Rubinstein et al., Machine Learning in Low-level Microarray Analysis, position paper in ACM SIGKDD Explorations (Special Issue on Microarray Data Mining), 5(2), December, 2003.

  • Statistical Algorithms Description Document, Whitepaper, Affymetrix Inc., Santa Clara (CA), 2002.

  • Liu et al., Analysis of high density expression microarrays with signed-rank call algorithms, Bioinformatics, 18(12), pgs 1593-1599, 2002.

  • Liu et al., Rank-based algorithms for analysis of microarrays, Proceedings of SPIE, Microarrays: Optical Technologies and Informatics, 4266, 2001.

mas5calls

mas5calls, a function of the affy package (starting with release 1.3) of the Bioconductor Project of bioinformatics tools for the R statistical computing environment, implements the Affymetrix Microarray Suite version 5.0 (MAS5) absolute detection call (of presence/marginal/absence of gene expression) algorithm with associated p-values, as described in the readings listed above. Further documentation and links will be added here shortly.

affyROCCHcomparison

affyROCCHcomparison is an online web tool that allows end-users of the Affymetrix GeneChip oligonucleotide microarray to quickly and easily
  1. compare available algorithms that detect gene expression (making presence/absence calls) and
  2. select necessary parameters for these algorithms (including the popular MAS5.0 p/a call)
based on the user's knowledge of
  1. the relative costs of making False Positive and False Negative misclassifications and
  2. the relative numbers of present and absent genes in the assayed sample
By utilizing this information on the utility of detection calls, users can make principled decisions on their choice of detection algorithm and algorithm parameters and be confident that the attained calls will be as useful as practically possible. Further documentation and links will be added here shortly.


[ top ]


Maintained by: Ben Rubinstein
Last updated: Fri Jul 27 22:33:29 PDT 2007