Web services accessed by ChimeraX fall into two general categories:
AlphaFold The AlphaFold tool Predict option and the command alphafold predict use Google Colab to run ColabFold, an open-source, optimized version of AlphaFold 2. Users will need to have a Google account and to sign into it via a browser.
BLAST Protein The Blast Protein tool performs protein sequence searches using a BLAST web service hosted by the UCSF Resource for Biocomputing, Visualization, and Informatics (RBVI). This service is also used by the Foldseek tool and the similarstructures blast command.
Clustal Omega Multiple sequence alignment with Clustal Omega can be performed using a web service hosted by the UCSF RBVI, accessed via the sequence align command or the Sequence Viewer context menu.
ESMFold The ESM Metagenomic Atlas provides a structure prediction server used by the command esmfold predict.
Foldseek The Foldseek tool and foldseek command perform a fast 3D structure similarity search using the Foldseek Search Service provided by the Söding and Steinegger groups.
Fast Sequence Database Search A K-mer sequence search service hosted by the UCSF RBVI is used by the commands alphafold match and esmfold match, as well as the AlphaFold and ESMFold Fetch capability. This search algorithm is fast but low-sensitivity, as appropriate for searching large sequence databases for matches with very high % identity.
MMseqs2 MMseqs2 is a very fast sequence similarity search method run by the Foldseek tool and the sequence search command using the RCSB web service.
Modeller Two tools run Modeller using a web service hosted by the UCSF RBVI:
- Modeller Comparative performs comparative (homology) modeling
- Model Loops performs untemplated modeling (loop remodeling and building missing segments)
MUSCLE Multiple sequence alignment with MUSCLE can be performed using a web service hosted by the UCSF RBVI, accessed via the sequence align command or the Sequence Viewer context menu.
SMILES Translator The SMILES translator provided by the National Cancer Institute CADD group is used to generate 3D molecule structures specified by SMILES string with the command open or in Build Structure.
Information can be retrieved from the following databases using the open command:
- RCSB Protein Data Bank – atomic structures, assemblies of atomic structures
- Protein Data Bank in Europe (PDBe) – atomic structures, assemblies of atomic structures, small-molecule connectivity and binding-site information
- Protein Data Bank Japan (PDBj) – atomic structures
- PDB Chemical Component Dictionary (CCD) – idealized structures of PDB chemical components (residue types)
- PDBe Electron Density Server (EDS) – (2Fo-Fc) electron density maps and (Fo-Fc) difference maps for some PDB entries
- Electron Microscopy Data Bank – electron density maps (ftp from wwpdb.org to .edu/.gov hosts, Chinese mirror site to .cn hosts, otherwise from ebi.ac.uk)
- PDB-Dev – integrative hybrid models
- AlphaFold Database – artificial intelligence-predicted protein structures
- AlphaMissense – predicted pathogenicity scores of mutations for analysis with mutationscores
- PubChem3D – modeled atomic coordinates for most entries in PubChem Compound
- UniProt – protein sequences, sequence feature annotations, predicted pathogenicity scores of variants