Tiger Team: tt-alphafold
Optimiziations for Alphafold
AlphaFold is a deep learning system that predicts a protein's 3D structure from its amino acid sequence. In 2021 an improved Version of Alphafold was released and installed on bwForCluster MLS&WISO for several scientific working groups. The performance of the software depends on fast access to databases and the use of GPUs. We want to explore different storage options for the databases and more powerful coprocessors to improve the performance and to enable more intensive use of the software.
Members of the Tiger-Teams:
ZMBH and BZH, Heidelberg University; HPC Competence Center MLS&WISO