Single-stranded DNA (ssDNA) binding proteins are crucial for biological processes including replication, transcription, and genomic integrity. We can better comprehend the function of the proteins and their interactions with ssDNA with structures of ssDNA binding proteins. However, the number of known ssDNA binding protein structures is still very small, despite recent advances in experimental protein determination methods. Lately, the implementation of AlphaFold2, a state-of-the art, artificial intelligence-based method, has led to a breakthrough in protein structure prediction. In this project we used the predicted ssDNA binding protein structures from AlphaFold2, on a dataset of annotated non-redundant ssDNA binding proteins. We investigated the quality of the predicted models and the relationship between protein size and model quality score.