Project Summary/Abstract Molecular interactions between proteins and single-stranded DNA (ssDNA) play crucial roles in many fundamental biological processes, such as DNA replication, DNA recombination, DNA repair, gene regulation, and genome stability. ssDNA-binding proteins encompass various degrees of binding specificity, ranging from highly specific to non-specific. Despite the critical roles of protein-ssDNA interactions in essential biological processes, the investigation of protein-ssDNA interactions lags far behind other types of protein-nucleic acids interactions, such as protein-double stranded DNA (dsDNA) interaction. Currently there are no comprehensive, well-annotated resources for studying protein- ssDNA interactions. This proposal aims to investigate the mechanism of protein-ssDNA interactions by first constructing well-annotated protein-ssDNA data resources and by performing comparative analysis at two different levels. The first is to compare structural features between protein-ssDNA complexes with different binding specificity. The second is to investigate the similarity/differences between protein- ssDNA and protein-dsDNA complexes. In addition, we propose to predict novel ssDNA binding proteins from the unannotated protein sequences by combining structure-based protein threading methods and sequence-based machine learning approaches.