An illumina paired-end and mate-pair short read simulator. This project attempts to model as many of the quirks that exist in Illumina data as possible. Some of these quirks include the potential for chimeric reads, and non-biotinylated fragment pull down in mate-pair libraries . Additionally the program provides the ability to model both site and base specific error, and scripts are provided to train this error model on real datasets. My hope in creating this program is to generate as realistic data as possible to assist in assessing the accuracy of genome assembly tools.