WAM: an improved algorithm for modelling antibodies on the WEB

Nicholas R.J. Whitelegg, Anthony R. Rees
2000 Protein Engineering Design & Selection  
except H3) frequently fall into one of between one and 10 canonical classes, a set for each CDR. Members of a canonical Centre for Protein Analysis and Design,University of Bath, Claverton class all have approximately the same backbone conformation. Down, Bath BA2 7AY, UK This is determined by the loop length and the presence of a of key residues, both in the CDR and in the framework arrees@compuserve.com (Foote and Winter, 1992), which hold the CDR in a given conformation by hydrogen bonding,
more » ... hydrogen bonding, electrostatic and/or hydro-An improved antibody modelling algorithm has been developed which incorporates significant improvements to phobic interactions. So, to model an unknown CDR, the sequence is examined, the appropriate canonical class assigned the earlier versions developed by Martin et al. (1989, 1991), Pedersen et al. (1992) and Rees et al. (1996) and known as and the most sequence-homologous known CDR used. For each loop except L2, a few examples fall outside existing AbM (Oxford Molecular). The new algorithm, WAM (for Web antibody modelling), has been launched as an online canonical classes, and, along with the H3 loop, must be modelled in other ways (see below). It is likely that further modelling service and is located at URL http://antibody.bath.ac.uk. Here we provide a summary only of the important canonical classes will be revealed as more crystal structures are solved, although to date no strictly 'canonical' classification features of WAM. Readers interested in further details are directed to the website, which gives extensive background has been possible for H3. The difficulty of modelling the H3 loop is due to the information on the methods employed. A brief description of the rationale behind some of the newer methodology extensive variability in both sequence and structure between different antibodies. There are essentially three approaches: (specifically, the knowledge-based screens) is also given. Keywords: antibody modelling algorithm/knowledge-based knowledge-based methods, such as database searching, where the closest matching database loop [either from antibodies or screens/WAM/WEB from the entire Protein Data Bank (PDB) (Berman et al., 2000)] in sequence and length is used as the model (e.g. as in ABGEN; Mandal et al., 1996) or ab initio methods, such as
doi:10.1093/protein/13.12.819 pmid:11239080 fatcat:pa3qxipn7rebzn6yy75gkmvji4