PhAST: Pharmacophore alignment search tool

Hähnke, Volker; Grgat, Tomislav; Proschak, Ewgenij; Steinhilber, Dieter; Schneider, Gisbert

doi:10.1002/jcc.21095

Cited by 25 publications

(22 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…47 Pharmacophore point graph features are computed similarly to molecular graph features, except that the molecular graph m = G m ( A m , B m ) is first mapped to an isomorphic graph p = G p ( A p , B p ), in which A p , B p are atoms and bonds with a small restricted set of labels. The labeling scheme, adapted from Hähnke, et al, 47 groups chemical motifs with similar reactivity. For example, all positively charged atoms are labeled the same, all negatively charged atoms are labeled the same, and all halides are labeled the same in the pharmacophore point graph.…”

Section: Machine Learning Stage 1: Reactive Site Filteringmentioning

confidence: 99%

Learning to Predict Chemical Reactions

Kayala

Azencott

Chen

et al. 2011

J. Chem. Inf. Model.

182

178

View full text Add to dashboard Cite

Being able to predict the course of arbitrary chemical reactions is essential to the theory and applications of organic chemistry. Approaches to the reaction prediction problems can be organized around three poles corresponding to: (1) physical laws; (2) rule-based expert systems; and (3) inductive machine learning. Previous approaches at these poles respectively are not high-throughput, are not generalizable or scalable, or lack sufficient data and structure to be implemented. We propose a new approach to reaction prediction utilizing elements from each pole. Using a physically inspired conceptualization, we describe single mechanistic reactions as interactions between coarse approximations of molecular orbitals (MOs) and use topological and physicochemical attributes as descriptors. Using an existing rule-based system (Reaction Explorer), we derive a restricted chemistry dataset consisting of 1630 full multi-step reactions with 2358 distinct starting materials and intermediates, associated with 2989 productive mechanistic steps and 6.14 million unproductive mechanistic steps. And from machine learning, we pose identifying productive mechanistic steps as a statistical ranking, information retrieval, problem: given a set of reactants and a description of conditions, learn a ranking model over potential filled-to-unfilled MO interactions such that the top ranked mechanistic steps yield the major products. The machine learning implementation follows a two-stage approach, in which we first train atom level reactivity filters to prune 94.00% of non-productive reactions with a 0.01% error rate. Then, we train an ensemble of ranking models on pairs of interacting MOs to learn a relative productivity function over mechanistic steps in a given system. Without the use of explicit transformation patterns, the ensemble perfectly ranks the productive mechanism at the top 89.05% of the time, rising to 99.86% of the time when the top four are considered. Furthermore, the system is generalizable, making reasonable predictions over reactants and conditions which the rule-based expert does not handle. A web interface to the machine learning based mechanistic reaction predictor is accessible through our chemoinformatics portal (http://cdb.ics.uci.edu) under the Toolkits section.

show abstract

Section: Machine Learning Stage 1: Reactive Site Filteringmentioning

confidence: 99%

Learning to Predict Chemical Reactions

Kayala

Azencott

Chen

et al. 2011

J. Chem. Inf. Model.

182

178

View full text Add to dashboard Cite

show abstract

“…20 We used weights 1, 2, 3, 4, 5, 10, 15, and 20, where a weight value of 1 corresponds to ''no weighting,'' in combination with the original PhAST scoring matrix (Table 3). 1 As an alternative, using explicit match-and mismatch-scores for each position in a query sequence would be possible with a position-specific scoring matrix. We chose implicit weighting factors as our aim was the extension of PhAST to position-specificity in a simple way.…”

Section: Weighted Phastmentioning

confidence: 99%

“…[1][2][3] It reduces each molecule to an unambiguous linear representation by describing its potential pharmacophore in three steps: (i) each nonhydrogen atom of the molecular graph is replaced by a potential pharmacophoric point (PPP) symbol, and hydrogen atoms are removed, (ii) vertices of this ''pharmacophore graph'' are canonically labeled, and iii) vertex symbols are concatenated into a string in increasing order according to their canonic labels. For virtual screening, both the screening compound collection (compound ''library'') and the query molecule(s) are converted as described, and the resulting ''PhASTsequences'' are compared using pairwise global sequence alignment.…”

Section: Introductionmentioning

confidence: 99%

Pharmacophore alignment search tool: Influence of scoring systems on text‐based similarity searching

Hähnke

Schneider

2011

J Comput Chem

Self Cite

View full text Add to dashboard Cite

The text-based similarity searching method Pharmacophore Alignment Search Tool is grounded on pairwise comparisons of potential pharmacophoric points between a query and screening compounds. The underlying scoring matrix is of critical importance for successful virtual screening and hit retrieval from large compound libraries. Here, we compare three conceptually different computational methods for systematic deduction of scoring matrices: assignment-based, alignment-based, and stochastic optimization. All three methods resulted in optimized pharmacophore scoring matrices with significantly superior retrospective performance in comparison with simplistic scoring schemes. Computer-generated similarity matrices of pharmacophoric features turned out to agree well with a manually constructed matrix. We introduce the concept of position-specific scoring to text-based similarity searching so that knowledge about specific ligand-receptor binding patterns can be included and demonstrate its benefit for hit retrieval. The approach was also used for automated pharmacophore elucidation in agonists of peroxisome proliferator activated receptor gamma, successfully identifying key interactions for receptor activation.

show abstract

“…2D FPT [51] 3DKeys [11a] ACCS-FP [52] CATS3D [53] Chem-X [54] FCFP [55] FEPOPS [20] Flexophore [56] GpiDAPH3 [57] Mtrees [58] OSPPREYS [59] PCFP [60] Pharm-IF [61] PharmPrint [62] PhAST [63] Screen/PMapper [64] SIFt [65] SQUID [66] Tuplets [67] UNITY2D [9a, 10]…”

Section: Fingerprint Basedmentioning

confidence: 99%

Pharmacophore Modeling Methods in Focused Library Selection – Applications in the Context of a New Classification Scheme

Luu¹,

Malcolm²,

Nadassy³

2011

CCHTS

View full text Add to dashboard Cite

A pharmacophore is a model which represents the key physico-chemical interactions that mediate biological activity. There is a long history of using pharmacophore modeling methods to select subsets of compounds, focused towards a specific target of interest. This paper will review existing computational methods for deriving and comparing pharmacophore models. We outline a new classification of pharmacophore methods based on the abstraction of the underlying chemical interactions which embody a pharmacophore, and the methods available to quantitatively compare them. Within the context of this classification, example studies, using specific pharmacophore modeling methods for focused library selection, will be discussed.

show abstract

PhAST: Pharmacophore alignment search tool

Cited by 25 publications

References 36 publications

Learning to Predict Chemical Reactions

Learning to Predict Chemical Reactions

Pharmacophore alignment search tool: Influence of scoring systems on text‐based similarity searching

Pharmacophore Modeling Methods in Focused Library Selection – Applications in the Context of a New Classification Scheme

Contact Info

Product

Resources

About