Progress Protein In Structure Assignment

Structure Assignenment Summer 1998 & Summer 1999

The pie-charts show the progress in structural assignment for the genome of Mycoplasma genitalium. A subset of the Mycoplasma genitalium genome that was structurally completely unassigned in early summer 1999 takining into account the results of several research groups (see, Teichmann SA, Chothia C, Gerstein M (1999), Curr. Opin. Struc. Biol. 9: 390-399. and , http://www.mrc-lmb.cam.ac.uk:80/genomes/MG/). We analysed this subset of the genome with PSI-BLAST and the most recent databases (sequence and structure). Two percentages of the residues of this subset could be assigned via close homology to sequence of known structure and 6 percentages to proteins of known structure via remote homology only. That means in about 1 years time there was a growth in structure assignment of 8 percentages. This is not only due to the growth of the structure database but also dependent on the growth of the protein sequence database because PSI-BLAST needs intermediate sequences to find remote homologies.

In an others years time one can expect more then 50% of the residues in the genome to be structurally annotated. Taking into account a much more rapid growth of the databases the assigned fraction will probably much larger.

The pie-charts below show the results of our structure assignment with databases from 1998 (left) and the 1999 (right). Fraction are calculated based on residues.

pie-chart of progress in structural assignments

Legend: in CC (coiled-coils), TM (Transmembrane helices), LC (low complexety regions), close (matches by close homologues), remote (matches remote homologues only), missing+new superfams (undetected homologues and new structural superfamilies)

get structural annotations for Mycoplasma genitalium ...