This program implements the multiple target similarity algorithms of Willett et al against a Daylight merlin pool.

The program operates in two major modes.

• The default is to read in a set of structures as SMILES and produce a consensus list of neighbours as described below.
• Alternatively in "turbo" mode a single structure is read in and a list of its nearest neighbours generated to act as multiple targets.
  • In this implementation, the similarity measure for this step is the same as for collection of the rest of the neighbours.

For each target, the pool is sorted by similarity and the user-defined number of neighbours retained.

These lists are fused by sorting them on either best rank or best similarity value against any target. Fusing in this way ensures that the original targets are retained in the final hit list.

Optional outputs are

• tdt for reading back into a client like xvmerlin
• SMILES with the sort key value as a 'name. This output can be piped into prado and a display program.
• tab delimited table with the following fields
  • Compound_ID
  • Best_rank
  • Best_similarity
  • Sum_of_ranks
  • Count_of_ranks
  • SMILES

Key literature references:

Whittle, M., Gillet, V.J., Willett, P., Alex, A. & Losel, J. "Enhancing the effectiveness of virtual screening by fusing nearest-neighbour lists: a comparison of similarity coefficients" Journal of Chemical Information and Computer Sciences 2004 44,1840-1848
Hert, J., Willett, P., Wilton, D.J., Acklin, P., Azzaoui, K., Jacoby, E. & Schuffenhauer, A. "Comparison of fingerprint-based methods for virtual screening using multiple bioactive reference structures" Journal of Chemical Information and Computer Sciences 2004 44, 1177-1185
Hert, J., Willett, P., Wilton, D.J., Acklin, P., Azzaoui, K., Jacoby, E. & Schuffenhauer, A. "Topological descriptors for similarity-based virtual screening using multiple bioactive reference structures" Organic and Biomolecular Chemistry 2004 2(22), 3256-3266

 Program multimer.

 Calling syntax: $ multimer [options] databasename

 Options:-
        -t <n>  the number of targets.(10)
        -f <n>  the maximum size of the final hitlist.(125)
        -n <n>  the number of neighbours to consider for each target.(50)
        Defaults are in parentheses
        -turbo  creates the target list by similarity to a single input SMILES
        Default is FALSE
        -ranks  create the output based on rank rather than value (default)
        -tdt    creates output as a file of SMILES in tdt format which will read
                directly into xvmerlin
        -smiles prints the space delimited SMILES followed by the best similarity
                value or rank
        Default is a tab delimited table:-
        Compound_ID Best_rank Best_similarity Sum_of_ranks Count_of_ranks SMILES

        -dos    produces output with cr/lf at the end of each data line to
                facilitate transfer to a Windows statistics package
        Default is FALSE i.e. unix output
 This program reads SMILES on stdin and returns the user requested output
 on stdout.

 Database specification format is: database%basepw@host:service:user%user-hostpw
 The database name must be specified (other fields are optional).