Skip to content

AlphaFold/ColabFold datasets

The datasets for AlphaFold, AlphaFold 3 and ColabFold are available on Midway3.

The datasets contain Multiple Sequence Alignment (MSA) data (Unifref90, MGnify, BFD, Uniclust30) for identifying homologous sequences that align with the input sequences and structure templates (PDB70 and PDB100) for predicting structures of the input sequences.

The paths to these datasets are as follows:

Path
AlphaFold 2 /software/alphafold-data-2.3
AlphaFold 3 /software/alphafold3.0-el8-x86_64/databases
ColabFold /software/colabfold-data

Please contact the RCC to request for updating the datasets when necessary. Example uses of these datasets with AlphaFold can be found in AlphaFold.