CompoundProtein Interaction Prediction Within Chemogenomics: Theoretical Concepts, Practical Usage, and Future Directions

Applicability Domain of Active Learning in Chemical Probe Identification: Convergence in Learning from Non-Specific Compounds and Decision Rule Clarification

Efficient identification of chemical probes for the manipulation and understanding of biological systems demands specificity for target proteins. Computational means to optimize candidate compound selection for experimental selectivity evaluation are being sought. The active learning virtual screening method has demonstrated the ability to efficiently converge on predictive models with reduced datasets, though its applicability domain to probe identification has yet to be determined. In this article, we challenge active learning’s ability to predict inhibitory bioactivity profiles of selective compounds when learning from chemogenomic features found in non-selective ligand-target pairs. Comparison of controls versus multiple molecule representations de-convolutes factors contributing to predictive capability. Experiments using the matrix metalloproteinase family demonstrate maximum probe bioactivity prediction achieved from only approximately 20% of non-probe bioactivity; this data volume is consistent with prior chemogenomic active learning studies despite the increased difficulty from chemical biology experimental settings used here. Feature weight analyses are combined with a custom visualization to unambiguously detail how active learning arrives at classification decisions, yielding clarified expectations for chemogenomic modeling. The results influence tactical decisions for computational probe design and discovery.

Virtual Screening Meets Deep Learning

BACKGROUND

Automated compound testing is currently the de facto standard method for drug screening, but it has not brought the great increase in the number of new drugs that was expected. Computer- aided compounds search, known as Virtual Screening, has shown the benefits to this field as a complement or even alternative to the robotic drug discovery. There are different methods and approaches to address this problem and most of them are often included in one of the main screening strategies. Machine learning, however, has established itself as a virtual screening methodology in its own right and it may grow in popularity with the new trends on artificial intelligence.

OBJECTIVE

This paper will attempt to provide a comprehensive and structured review that collects the most important proposals made so far in this area of research. Particular attention is given to some recent developments carried out in the machine learning field: the deep learning approach, which is pointed out as a future key player in the virtual screening landscape.

The Future of Computational Chemogenomics

Following the elucidation of the human genome, chemogenomics emerged in the beginning of the twenty-first century as an interdisciplinary research field with the aim to accelerate target and drug discovery by making best usage of the genomic data and the data linkable to it. What started as a systematization approach within protein target families now encompasses all types of chemical compounds and gene products. A key objective of chemogenomics is the establishment, extension, analysis, and prediction of a comprehensive SAR matrix which by application will enable further systematization in drug discovery. Herein we outline future perspectives of chemogenomics including the extension to new molecular modalities, or the potential extension beyond the pharma to the agro and nutrition sectors, and the importance for environmental protection. The focus is on computational sciences with potential applications for compound library design, virtual screening, hit assessment, analysis of phenotypic screens, lead finding and optimization, and systems biology-based prediction of toxicology and translational research.

Selection of Informative Examples in Chemogenomic Datasets

High-throughput and high-content screening campaigns have resulted in the creation of large chemogenomic matrices. These matrices form the training data which is used to build ligand-target interaction models for pharmacological and chemical biology research. While academic, government, and industrial efforts continuously add to the ligand-target data pairs available for modeling, major research efforts are devoted to improving machine learning techniques to cope with the sparseness, heterogeneity, and size of available datasets as well as inherent noise and bias. This "race of arms" has led to the creation of algorithms to generate highly complex models with high prediction performance at the cost of training efficiency as well as interpretability.In contrast, recent studies have challenged the necessity for "big data" in chemogenomic modeling and found that models built on larger numbers of examples do not necessarily result in better predictive abilities. Automated adaptive selection of the training data (ligand-target instances) used for model creation can result in considerably smaller training sets that retain prediction performance on par with training using hundreds of thousands of data points. In this chapter, we describe the protocols used for one such iterative chemogenomic selection technique, including model construction and update as well as possible techniques for evaluations of constructed models and analysis of the iterative model construction.

Impact of Molecular Descriptors on Computational Models

Molecular descriptors encode a wide variety of molecular information and have become the support of many contemporary chemoinformatic and bioinformatic applications. They grasp specific molecular features (e.g., geometry, shape, pharmacophores, or atomic properties) and directly affect computational models, in terms of outcome, performance, and applicability. This chapter aims to illustrate the impact of different molecular descriptors on the structural information captured and on the perceived chemical similarity among molecules. After introducing the fundamental concepts of molecular descriptor theory and application, a step-by-step retrospective virtual screening procedure guides users through the fundamental processing steps and discusses the impact of different types of molecular descriptors.

Exome Sequencing Landscape Analysis in Ovarian Clear Cell Carcinoma Shed Light on Key Chromosomal Regions and Mutation Gene Networks

Previous studies have reported genome-wide mutation profile analyses in ovarian clear cell carcinomas (OCCCs). This study aims to identify specific novel molecular alterations by combined analyses of somatic mutation and copy number variation. We performed whole exome sequencing of 39 OCCC samples with 16 matching blood tissue samples. Four hundred twenty-six genes had recurrent somatic mutations. Among the 39 samples, ARID1A (62%) and PIK3CA (51%) were frequently mutated, as were genes such as KRAS (10%), PPP2R1A (10%), and PTEN (5%), that have been reported in previous OCCC studies. We also detected mutations in MLL3 (15%), ARID1B (10%), and PIK3R1 (8%), which are associations not previously reported. Gene interaction analysis and functional assessment revealed that mutated genes were clustered into groups pertaining to chromatin remodeling, cell proliferation, DNA repair and cell cycle checkpointing, and cytoskeletal organization. Copy number variation analysis identified frequent amplification in chr8q (64%), chr20q (54%), and chr17q (46%) loci as well as deletion in chr19p (41%), chr13q (28%), chr9q (21%), and chr18q (21%) loci. Integration of the analyses uncovered that frequently mutated or amplified/deleted genes were involved in the KRAS/phosphatidylinositol 3-kinase (82%) and MYC/retinoblastoma (75%) pathways as well as the critical chromatin remodeling complex switch/sucrose nonfermentable (85%). The individual and integrated analyses contribute details about the OCCC genomic landscape, which could lead to enhanced diagnostics and therapeutic options.

Novel Method Proposing Chemical Structures with Desirable Profile of Activities Based on Chemical and Protein Spaces

Active molecules among numerous chemical structures in a chemical database can be searched easily by statistical prediction of compound-protein interactions. However, constructing a simple prediction model against one protein does not aid drug design, because detecting chemical structures that act similarly against multiple proteins is necessary for preventing side effects of the potential drug. To tackle this problem, we propose a new method that visualizes chemical and protein spaces. For simultaneous visualization of both spaces, we employ a counterpropagation neural network (CPNN) and develop a new visualization method named multi-input CPNN (MICPNN). In a case study of the kinase protein family, the MICPNN model predicted accurately the complex relationships between compounds and proteins. The proposed method identified chemical structures with promising activity against kinases. Our proposed method is also applicable to other protein families, such as G-protein coupled receptors, ion channels and transporters.

Active learning for computational chemogenomics

Aim: Computational chemogenomics models the compound–protein interaction space, typically for drug discovery, where existing methods predominantly either incorporate increasing numbers of bioactivity samples or focus on specific subfamilies of proteins and ligands. As an alternative to modeling entire large datasets at once, active learning adaptively incorporates a minimum of informative examples for modeling, yielding compact but high quality models. Results/methodology: We assessed active learning for protein/target family-wide chemogenomic modeling by replicate experiment. Results demonstrate that small yet highly predictive models can be extracted from only 10–25% of large bioactivity datasets, irrespective of molecule descriptors used. Conclusion: Chemogenomic active learning identifies small subsets of ligand–target interactions in a large screening database that lead to knowledge discovery and highly predictive models.

Data Mining of Chemogenomics Data Using Bi-Modal PLS Methods and Chemical Interpretation for Molecular Design

Chemogenomics is a new strategy in drug discovery for interrogating all molecules capable of interacting with all biological targets. Because of the almost infinite number of drug-like organic molecules, bench-based experimental chemogenomics methods are not generally feasible. Several in silico chemogenomics models have therefore been developed for high-throughput screening of large numbers of drug candidate compounds and target proteins. In previous studies, we described two novel bi-modal PLS approaches. These methods provide a significant advantage in that they enable direct connections to be made between biological activities and ligand and protein descriptors. In this special issue, we review these two PLS-based approaches using two different chemogenomics datasets for illustration. We then compare the predictive and interpretive performance of the two methods using the same congeneric data set.

Computational chemogenomics: is it more than inductive transfer?

High-throughput assays challenge us to extract knowledge from multi-ligand, multi-target activity data. In QSAR, weights are statically fitted to each ligand descriptor with respect to a single endpoint or target. However, computational chemogenomics (CG) has demonstrated benefits of learning from entire grids of data at once, rather than building target-specific QSARs. A possible reason for this is the emergence of inductive knowledge transfer (IT) between targets, providing statistical robustness to the model, with no assumption about the structure of the targets. Relevant protein descriptors in CG should allow one to learn how to dynamically adjust ligand attribute weights with respect to protein structure. Hence, models built through explicit learning (EL) by including protein information, while benefitting from IT enhancement, should provide additional predictive capability, notably for protein deorphanization. This interplay between IT and EL in CG modeling is not sufficiently studied. While IT is likely to occur irrespective of the injected target information, it is not clear whether and when boosting due to EL may occur. EL is only possible if protein description is appropriate to the target set under investigation. The key issue here is the search for evidence of genuine EL exceeding expectations based on pure IT. We explore the problem in the context of Support Vector Regression, using more than 9,400 pKi values of 31 GPCRs, where compound-protein interactions are represented by the concatenation of vectorial descriptions of compounds and proteins. This provides a unified framework to generate both IT-enhanced and potentially EL-enabled models, where the difference is toggled by supplied protein information. For EL-enabled models, protein information includes genuine protein descriptors such as typical sequence-based terms, but also the experimentally determined affinity cross-correlation fingerprints. These latter benchmark the expected behavior of a quasi-ideal descriptor capturing the actual functional protein-protein relatedness, and therefore thought to be the most likely to enable EL. EL- and IT-based methods were benchmarked alongside classical QSAR, with respect to cross-validation and deorphanization challenges. A rational method for projecting benchmarked methodologies into a strategy space is given, in the aims that the projection will provide directions for the types of molecule designs possible using a given methodology. While EL-enabled strategies outperform classical QSARs and favorably compare to similar published results, they are, in all respects evaluated herein, not strongly distinguished from IT-enhanced models. Moreover, EL-enabled strategies failed to prove superior in deorphanization challenges. Therefore, this paper raises caution that, contrary to common belief and intuitive expectation, the benefits of chemogenomics models over classical QSAR are quite possibly due less to the injection of protein-related information, and rather impacted more by the effect of inductive transfer, due to simultaneous learning from all of the modeled endpoints. These results show that the field of protein descriptor research needs further improvements to truly realize the expected benefit of EL.