Aggregator Advisor

From DISI
(Redirected from Aggregator advisor)
Jump to navigation Jump to search

About

Aggregator advisor is a service to predict aggregation based on similarity to know aggregators, and calculated LogP.

Aggregator Advisor is a system for finding similar molecules to the query that are known aggregators or non-aggregators. The heart of the system is the Chemaxon molecular query system. The Aggregate Advisor is a wrapper around the chaemaxon library. The operation sequence of Aggregate Advisor is following

Cite Aggregator Advisor

To cite Aggregator Advisor, please use: Irwin, Duan, Torosyan, Doak, Ziebart, Sterling, Tumanian and Shoichet, J Med Chem, 2015, 58(1 7), 7076-7087[1].

Query procedure

For a query molecule(submitted through the web-interface):

1. Look up molecule

  If (aggregator) then
  say so. give literature reference (justification)
  stop
  else (ambiguous) then
  say so. give literature reference (justification)
  stop
  else (non-aggregator) then
  say so. give literature reference (justification)
  continue
  endif

2. Compute similarity to aggregators, ambiguous,

  If (similar to aggregator) then
  say so. give literature reference (justification),
  similar molecule and similarity level
  stop
  else (similar to ambiguous) then
   say so. give literature reference (justification)
  similar molecule and similarity level
  endif

3. Compute murcko scaffold.

check similarity of murcko scaffold vs aggregators, ambiguous

  If (murcko is substructure of aggregator) th
  say so. give literature reference (justification)
  show database molecule we are a substructure of
  stop
  else (murcko is substructure of ambiguous) then
  say so. give literature reference (justification)
  show database molecule we are a substructure of
  endif

4. Compute physical properties.

  if logP > 3.5, say: warning  
  else 
  For now, say:  no reason to suspect this is an aggregators.

System design

The Aggregate Advisor application consists of several software layers. each layer wraps around the lower one. The top layer is the UI engine, which allows a user of the system to draw and submit a query molecule and view the results in a formatted form, together with the structural image of the response molecules. Additionally the user of te system can browse the library of known aggregators. This level is built in PHP using CodeIgniter Library. For drawing capability it uses the WebME editor. Adittionally it requests the SEA rendering system for displaying the images of the molecules.

The next layer is the web-service, which allows to query the AggregatorAdvisor. The Front-end uses the aggregate advisor web-service the same way as any third-party would use it. The users of the system make XML-RPC requests and receive the query responses in XML-RPC format. The protocol of operation is given in the section *Web Services Description

References