The GeneBench server provides a valuable service for creation of your own non-homologous gene dataset. Users are allowed to input their nucleotide dataset along with their CDS annotation in required format. The GeneBench server extracts the gene information of the sequences from the annotation and translates the corresponding genes into protein products. The Proteins are then filtered for minimum homology allowed between protein sequences using the PROSET program. PROSET will perform a pairwise comparison on all pairs of proteins from the dataset which have at least one defined -mer in common. Of each such pair the smaller protein is eliminated from the list if the two sequences have an overall block identity of at least minimum defined % identity. After the removal of such sequences that have similarity above the required threshold, the new set of non-homologous sequences are generated. Users are intimated by email to download their sequences within 1 day of completion of process after which the dataset will be removed from the server. Confidentiality of the users and security of their datasets is maintained for all processes. Due to limitation of PROSET program to handle large datasets, a limit of 4000 gene sequences is applied to each user with the total length of their protein product adding upto 2 Million or 2000000 amino acids.