ABCMdb Usage

Leaving comments and annotations
Homologues and alignments
Protein structures
File formats


The ABCMdb web interface has been designed to be self-explanatory and easily explorable. Most colored texts are clickable and point to a page with detailed information on the highlighted term. To browse the mutations of a specific protein, start from the Browse menu, expand the protein family of interest with the "show" button, and click on any of the "list mutations" links. This will take you to the protein details view, where clicking on the "details" link will bring up the detailed mutation view of the corresponding mutation.

Mutations can be displayed in three different views. The default view is the detailed mutation view, which displays information on all mentions of a specific mutation of a protein. Here you can use the show/hide buttons to navigate to the information of interest. The "switch to compact view" link at the top of the detailed view takes you to the compact mutation view, which displays essentially the same information, but with less detail on the publications, and all sentence panels are unrolled. Clicking on the PMID link in any view takes you to the publication view, where all mutations found in a specific paper are displayed.

External links to other databases are provided on some views. The "reference sequence" link on the browse and protein details pages shows the sequence of the corresponding protein that was used during text mining, and a link to the UniProt record of the sequence is also provided. Clicking on the PubMed link next to a publication takes you to the PubMed record of the corresponding paper.

Leaving comments and annotations

Registered users will find occasional text boxes where they can comment on mutations, publications, and sentences. Comments can be submitted by pressing Enter, if you want to start a new line in your comment, press the Shift+Enter key combination. Not all items can be commented in all views, comments for mutations can be submitted in the detailed and compact mutation views, publications can only be commented in the detailed mutation view, while sentences can be commented in any of the mutation views, and the publication view.

Because of automatic extraction, erroneously reported mutations might appear in the database. Therefore, the manual verification of mutation mentions is a valuable contribution to enhance the quality of the data in the ABCMdb. Registered users can change the status of individual mutation mentions in a pop-up box that appears when the highlighted mutation mention is clicked. We encourage registered users to annotate the dataset using this method in case they spot an error. Mutation mentions are currently categorized under several labels as explained in the table below.

NEWThis instance of the mutation is yet unclassified.
VERIFIEDThis instance is verified to indicate a proper mutation of the associated protein.
PROTEIN_MISMATCHThis instance is incorrectly associated with the protein.
NUCLEOTIDEThis instance represents a nucleotide change rather than an amino acid change.
ERRONEOUSThis instance is erroneous for other reasons.

We encourage all database users to register and leave comments and annotations on the data to increase the quality and reliability of the database.

Homologues and alignments

In the protein details view, clicking on the homologous mutations link next to a mutation will allow you to select an alignment for homology analysis. After selecting an alignment, the search homologues view is displayed, which shows mutations in other ABC proteins that are in homologous positions to the selected mutation. The page will also show mutations in other ABC proteins that are within 5 residues in sequence of the homologous position. Clicking on "details" next to a mutation will take you to the detailed mutation view of the selected mutation. If only a single alignment is present for the selected protein, the alignment selection page will not appear, and you will be taken directly to the search homologues view.

The segment of the alignment provided at the bottom of the search homologues view is centered at the position whose homologous residues are displayed, and clicking on any of the red residues will show a list of mutations of that residue.

In order to use these features, multiple alignments of the proteins must be available. Various alignments of NBDs and full length transporters are provided by the web site, usually one for each family, for proteins from multiple families with similar topology, and all ABC proteins. In addition, custom alignments can be uploaded by registered users under the MyProfile menu that can be used with these functions.

Protein structures

To display mutated residues on a 3D view of a protein structure, click on the map mutations link in the protein details view next to a protein name. The web page will first prompt you to select an alignment if more than one alignment is available for the selected protein. Then, you will be prompted to select a structure to use for visualization, if more than one is available. After selecting an alignment and a structure the map mutations view displays an interactive 3D view of the selected protein structure. The bottom of the page will display the selected alignment. Clicking on any letter in the top (underlined) row of the alignment will highlight the selected residue in the 3D display. Clicking on the red residues in the other rows of the alignment will display a list of mutations for that protein and residue.

Currently, the following structural models are provided by the web site.

ProteinName of modelDescription
ABCB1P-gp_holo_WieseHuman P-gp homology model in the bottom-closed state based on Sav1866 as in Globisch et al. 2008, ChemMedChem 3:280-95. [PubMed]
ABCB1hMDR1_apohMDR1 homology model based on 3G5U.
ABCC1MRP1_model_ColeHuman MRP1 full length model in the bottom-closed state as in DeGorter et al. 2008, Biochem Biophys Res Commun 365:29-34. [PubMed]
ABCC6MRP6_model_VaradiHuman MRP6 full length model in the bottom-closed state as in Fülöp et al. 2009, Biochem Biophys Res Commun 379:706-9. [PubMed]
ABCC72BBOPDB structure, "Human NBD1 with Phe508".
ABCC72BBSPDB structure, "Human deltaF508 NBD1 with three solubilizing mutations".
ABCC72BBTPDB structure, "Human deltaF508 NBD1 with two solublizing mutations".
ABCC72PZEPDB structure, "Minimal human CFTR first nucleotide binding domain as a head-to-tail dimer".
ABCC72PZFPDB structure, "Minimal human CFTR first nucleotide binding domain as a head-to-tail dimer with delta F508".
ABCC72PZGPDB structure, "Minimal human CFTR first nucleotide binding domain as a monomer".
ABCG2ABCG2_NBDHuman ABCG2 NBD homology model based on 1Q12.

Custom protein structures can be uploaded by registered users under the MyProfile menu item, and used for 3D visualization.

File formats

Structural files can be uploaded in PDB format.

Uploaded alignment files must be in ClustalW format, where the protein names are in ABCNN_xxx format. NN should denote the family and canonical number of the protein, e.g. G2 or B11, and xxx should be the starting position of the sequence. For example, a protein name of ABCC1_673 indicates that the protein sequence belongs to the human ABCC1 protein, and that the first character in the sequence is at position 673 of the protein.

Do not forget to check the validity of residue numbering in the alignments and structures you upload!