PowerTrim: An automated decision support algorithm for preprocessing family-based genetic data

Statistical genetics software packages for linkage analysis have their own unique constraints on the size and shape of the pedigrees they can process. As a result, researchers are often forced to exclude from analysis some individuals in a given family. Existing procedures for reducing pedigree size to fit computational constraints use arbitrary rules and are not interactive. However, judicious evaluation of which subject(s) to remove to minimize loss of information involves consideration of many factors, including informativeness owing to position in pedigree, availability of genotypic information, and quality of phenotypic information. Thus, automation of this task would be of significant benefit. We designed an interactive algorithm (PowerTrim) that provides the user access to detailed information with which to make informed decisions. In addition, PowerTrim checks for transcriptional and data-entry errors, which can be very time-consuming to localize manually.