Computational analyses of mechanism of action (MoA): data, methods and integration

Abstract

The elucidation of a compound's Mechanism of Action (MoA) is a challenging task in the drug discovery process, but it is important in order to rationalise phenotypic findings and to anticipate potential side-effects. Bioinformatic approaches, advances in machine learning techniques and the increasing deposition of high-throughput data in public databases have significantly contributed to recent advances in the field, but it is not straightforward to decide which data and methods are most suitable to use in a given case. In this review, we focus on these methods and data and their applications in generating MoA hypotheses for subsequent experimental validation. We discuss compound-specific data such as -omics, cell morphology and bioactivity data, as well as commonly used supplementary prior knowledge such as network and pathway data, and provide information on databases where this data can be accessed. In terms of methodologies, we discuss both well-established methods (connectivity mapping, pathway enrichment) as well as more developing methods (neural networks and multi-omics integration). Finally, we review case studies where the MoA of a compound was successfully suggested from computational analysis by incorporating multiple data modalities and/or methodologies. Our aim for this review is to provide researchers with insights into the benefits and drawbacks of both the data and methods in terms of level of understanding, biases and interpretation - and to highlight future avenues of investigation which we foresee will improve the field of MoA elucidation, including greater public access to -omics data and methodologies which are capable of data integration.

Fig. 1. Overview of the different types of data/information used in MoA studies and the various levels that MoA can be defined on, as reviewed in this paper. This includes experimental data, such as transcriptomics data, and data resources which are used to provide biological context to experimental data, such as pathway and network data. Created with BioRender.

Fig. 2. Schematic description of the cell painting assay demonstrated with the Warfarin compound. Created with BioRender using cell images from the Image Data Resource (IDR0036).

Fig. 3. The merged mTOR signalling pathway from KEGG (blue), Reactome (orange) and Wikipathways (green) visualised in PathME viewer. The intersection sizes represent the number of entities in common vs. the number of entities in each pathway. We observe that, for the same pathway, the information from 3 different sources varies. Visualisation created with PathMe Viewer.

Fig. 4. Connectivity map procedure (adapted from original article). (A) The biological state of interest should be represented as a gene expression signature (query), from which the top up- and down-regulated genes are interrogated. (B) The query signature is compared against reference profiles to compute connectivity. (C) The reference profiles are ranked in terms of both magnitude and direction (positive or negative) of connectivity to the query signature.

Fig. 5. The GO hierarchy is skewed, and contains redundant terms. Tools such as GOATOOLS can be used to correct for the skewed nature of GO ontology. Here, three terms (A, B and S) have the same level of hierarchy but different descendants, which illustrates the complexity of using GO terms for enrichment analysis. Figure adapted from Klopfenstein et al. with permission from the authors, copyright 2018.

Fig. 6. (A) Demonstration of model overview. Multi Omics Factor Analysis (MOFA) takes a number of data matrices as input from different data modalities and decomposes these matrices into a matrix of factors for each sample and weight matrices, one for each data modality. (B) Downstream analysis of MOFA model including variance decomposition, assessing the proportion of variance explained by each factor in each data modality, inspection of factors and imputation of missing values. Created with BioRender.

Maria-Anna Trapotsi

Layla Hosseini-Gerami

Andreas Bender