Advances in Computational Medicinal Chemistry: Matched Molecular Pair Analysis

Abstract

Preclinical Research Herein we discuss the concept of matched molecular pairs (MMPs) and relevant medicinal chemistry applications. In recent years, this concept has become popular in medicinal chemistry as it provides a formal and general basis to establish structural relationships between compounds, identify chemical changes that transform structurally related compounds into each other, and facilitate a large‐scale analysis of structure–activity relationships. Algorithms have been developed for the systematic computational identification of qualifying pairs in large compound databases. In addition to a number of other areas such as graphical analysis and visualization of structure–activity relationships, systematic MMP analysis has significantly advanced the field of computational medicinal chemistry, providing a basis for many practical applications. Characteristic features of this concept include that it eliminates subjective criteria from structural comparisons, alleviates the need to employ predefined molecular hierarchies in studying structural relationships, and elucidates distinguishing structural features between active compounds in a highly intuitive manner. Hence, the results of computational MMP analysis are readily accessible to medicinal chemists, which makes this type of analysis highly attractive for a variety of applications.

Automatic Identification of Analogue Series from Large Compound Data Sets: Methods and Applications

“…The expression was coined by Kenny and Sadowski [ 24 ], who considered pairs of compounds related to each other by a single (predefined) transformation [ 25 ]. Many current approaches define an MMP as a pair of compounds related to each other by a small structural change at a single site (see Figure 1 ).The traceability provided by limited structural differences makes the SAR insights obtained through MMP analysis (MMPA) more intuitive than those produced by many other similarity-based analyses [ 26 ]. For example, MMPA can systematically capture the chemical knowledge in a database and quantify the average effect of a given transformation [ 27 , 28 , 29 ].…”

“…Table 1 summarizes these approaches. For an extensive review of MMP algorithms, see Reference [ 26 ] or Reference [ 28 ].…”

Matched Molecular Pair Analysis in Short: Algorithms, Applications and Limitations

“…These are discussed in various publications [12], [38] and include: the problem of generalized relationships like global versus local effects, molecular context, database, time and data set dependencies and definition of the relevance of observed difference (e.g. cut-off at 2 fold change).…”

“…Among others these include: Drug-Guru [6], [7], Buy me Grease [7], WizePairZ [8], T-Analyse and T-Morph [9], VAMMPIRE [10] as well as the Hussain-Rea MMP algorithm [11] (Table 1). The MMP concept has been further developed into Matched Pair Series [12], [13] or Matched Molecular Series (MMS) [14] to describe a set of compounds (not only a pair) differing by only a single chemical transformation.…”

What compound should I make next? Using Matched Molecular Series for prospective medicinal chemistry

“…A Matched Molecular Pair (MMP) is a pair of compounds which differ only by a well-defined structural transformation [ 1 , 2 ]. Together with large-scale mining of activity or physicochemical data, matched molecular pair analysis (MMPA) has the potential to aid the design of molecules with improved properties by highlighting favourable transformations.…”