This tool takes a normalized dataset, and combines all probes that have been mapped to the same gene name into a single expression estimate for that gene. Probes are averaged together to get a single estimate of expression.
By default, gene names are used as identifiers after combination. This can be changed by adjusting the parameter to the setting 'yes'. Using this setting, every gene is identified by one of the original identifiers. After that, all normal analyses requiring annotation information can be used (not so with gene names).
A new, smaller dataset