The use of certain types of microbes for the treatment of radioactive waste is determined by the surprising capacity of these organisms for adaptation to ionizing radiation and desiccation. These microbes have the ability to bioremediation of radioactive mixed waste and especially to detoxify mercury and toluene in the waste. In recent years, in vitro studies have investigated the causes of the high resistance of some organisms to ionizing radiation to improve the treatment of radioactive waste. Bioinformatics analysis (in silico) may lead in vitro experiments, and then save time and money. The main objective of this project is to design and develop a decision system for the prediction of the phenotype of bacteria over the radioresistance (radioresistant or radiosensitive). This system has two main components. The first is a database containing various information (genomics, publications, etc.) about these bacteria. The second component is a system for predicting the radioresistance of new studied bacteria. Genomic knowledge related to these bacteria must be used for bioremediation of radioactive waste.

• MIL-Align, a MIL approach for predicting IRRB using proteins implicated in basal DNA repair in IRRB. We used a local alignment technique to measure the similarity between protein sequences of the studied bacteria to predict ionizing-radiation-resistant bacteria.
  
• IRRBMotif, a motif-based approach for predicting IRRB. We used a set of discriminative protein motifs to construct a set of machine learning classifiers that predict whether a bacterium belongs to IRRB or IRSB.