DNA hybridization methods have become the most widely used tools in molecular biology to identify organisms and evaluate gene expression levels. PCR (Polymerase Chain Reaction)-based methods, fluorescent in situ hybridization (FISH) and the recent development of DNA microarrays as a high throughput technology need efficient primers or probes design. Evaluation of the metabolic capacities of complex microbial communities found in terrestrial or aquatic environments requires new probe design algorithms that reflect the genetic diversity. As only a small part of the microbial diversity is known, gene sequences deposited in international databases do not reflect the entire diversity. In this context we propose to use oligopeptide sequences for the design of complete set of DNA probes that are able to target the entire genetic diversity of genes encoding enzymes. Due to the degenerated genetic code backtranslation must be managed efficiently. To our knowledge no software has been developed to propose a full backtranslation. This complexity is tractable since we only need to focus on short oligopeptides for DNA probe design. We propose new algorithms that perform a high performance oligopeptide backtranslation into all potential nucleic sequences. We use different efficient techniques such as memory mapping to perform such a computing. We also propose a MPI parallel computing that reduces the whole execution time using data load balancing and network file stream distribution on a cluster architecture.
