The string often appears in the context of deep learning research involving large protein databases. It specifically refers to the 48,225 entries in the Multi-domain Protein structure Database (MPDB) used to train various deep learning models for protein structure prediction and domain assembly.
: This study clustered the 48,225 proteins from the MPDB to create a non-redundant training set for predicting rigid motion in protein structures. 48225 rar
: This paper introduces DeepAssembly , a protocol that uses deep learning to predict inter-domain interactions. The string often appears in the context of
Key research papers that reference this specific dataset size include: : This paper introduces DeepAssembly , a protocol
: This paper discusses improving protein modeling accuracy using a database of 48,225 multi-domain proteins for analogue detection.
Additionally, the number appears as a page index or identifier in the Proceedings of the 41st International Conference on Machine Learning (ICML 2024) for the paper titled "Ranking-based Client Imitation Selection for Efficient Federated Learning".