Randomly generates reads to aid in the testing of alignment programs. Reads of a specified length are made from randomly reading chunks from the input fasta sequence. The number of reads made is determined by specifying the average coverage of the input sequence. User-specified SNP error rates can simulate sequence error. To simulate heterozygous/polyploid reads, supply multiple sequences in the input file; the script will make reads for each sequence, writing them all to the same file. The script can also simulate paired-end reads with a specified range of insert sizes.