FP-ADMET: a compendium of fingerprint-based ADMET prediction models

Motivation The absorption, distribution, metabolism, excretion, and toxicity (ADMET) of drugs plays a key role in determining which among the potential candidates are to be prioritized. In silico approaches based on machine learning methods are becoming increasing popular, but are nonetheless limited by the availability of data. With a view to making both data and models available to the scientific community, we have developed FPADMET which is a repository of molecular fingerprint-based predictive models for ADMET properties. Summary In this article, we have examined the efficacy of fingerprint-based machine learning models for a large number of ADMET-related properties. The predictive ability of a set of 20 different binary fingerprints (based on substructure keys, atom pairs, local path environments, as well as custom fingerprints such as all-shortest paths) for over 50 ADMET and ADMET-related endpoints have been evaluated as part of the study. We find that for a majority of the properties, fingerprint-based random forest models yield comparable or better performance compared with traditional 2D/3D molecular descriptors. Availability The models are made available as part of open access software that can be downloaded from https://gitlab.com/vishsoft/fpadmet.

Convolutional neural networks (CNNs): concepts and applications in pharmacogenomics

Convolutional neural networks (CNNs) have been used to extract information from various datasets of different dimensions. This approach has led to accurate interpretations in several subfields of biological research, like pharmacogenomics, addressing issues previously faced by other computational methods. With the rising attention for personalized and precision medicine, scientists and clinicians have now turned to artificial intelligence systems to provide them with solutions for therapeutics development. CNNs have already provided valuable insights into biological data transformation. Due to the rise of interest in precision and personalized medicine, in this review, we have provided a brief overview of the possibilities of implementing CNNs as an effective tool for analyzing one-dimensional biological data, such as nucleotide and protein sequences, as well as small molecular data, e.g., simplified molecular-input line-entry specification, InChI, binary fingerprints, etc., to categorize the models based on their objective and also highlight various challenges. The review is organized into specific research domains that participate in pharmacogenomics for a more comprehensive understanding. Furthermore, the future intentions of deep learning are outlined.

Extended similarity indices: the benefits of comparing more than two objects simultaneously. Part 1: Theory and characteristics†

Quantification of the similarity of objects is a key concept in many areas of computational science. This includes cheminformatics, where molecular similarity is usually quantified based on binary fingerprints. While there is a wide selection of available molecular representations and similarity metrics, there were no previous efforts to extend the computational framework of similarity calculations to the simultaneous comparison of more than two objects (molecules) at the same time. The present study bridges this gap, by introducing a straightforward computational framework for comparing multiple objects at the same time and providing extended formulas for as many similarity metrics as possible. In the binary case (i.e. when comparing two molecules pairwise) these are naturally reduced to their well-known formulas. We provide a detailed analysis on the effects of various parameters on the similarity values calculated by the extended formulas. The extended similarity indices are entirely general and do not depend on the fingerprints used. Two types of variance analysis (ANOVA) help to understand the main features of the indices: (i) ANOVA of mean similarity indices; (ii) ANOVA of sum of ranking differences (SRD). Practical aspects and applications of the extended similarity indices are detailed in the accompanying paper: Miranda-Quintana et al. J Cheminform. 2021. 10.1186/s13321-021-00504-4. Python code for calculating the extended similarity metrics is freely available at: https://github.com/ramirandaq/MultipleComparisons.

Clandestine nanoelectromechanical tags for identification and authentication

The realization of truly unclonable identification and authentication tags is the key factor in protecting the global economy from an ever-increasing number of counterfeit attacks. Here, we report on the demonstration of nanoscale tags that exploit the electromechanical spectral signature as a fingerprint that is characterized by inherent randomness in fabrication processing. Benefiting from their ultraminiaturized size and transparent constituents, these clandestine nanoelectromechanical tags provide substantial immunity to physical tampering and cloning. Adaptive algorithms are developed for digital translation of the spectral signature into binary fingerprints. A large set of tags fabricated in the same batch is used to estimate the entropy of the corresponding fingerprints with high accuracy. The tags are also examined under repetitive measurements and temperature variations to verify the consistency of the fingerprints. These experiments highlight the potential of clandestine nanoelectromechanical tags for the realization of secure identification and authentication methodologies applicable to a wide range of products and consumer goods.

Fingerprint-based 2D-QSAR Models for Predicting Bcl-2 Inhibitors Affinity

Background: Bcl-2 family plays an essential role in the cell cycle events incorporating survival, proliferation, and differentiation in normal and neoplastic neuronal cells. Thus, it has been validated as a principal target for the treatment of cancer. For this reason, we will build a model based on a large number of Bcl-2 inhibitors to predict the activities of new compounds as future Bcl-2 inhibitors. Methods: In this study, QSAR models were successfully used to predict the inhibitory activity against Bcl-2 for a set of compounds collected from BDB (Binding database). The kPLS (kernelbased Partial Least-Square) method implemented in Schrodinger's Canvas, was used for searching the correlation between pIC50 and binary fingerprints for a set of known Bcl-2 inhibitors. Results and Discussion: Models based on binary fingerprints with two kPLS factors have been found with decent predictive power (q2 > 0.58), while the optimal number of factors is about 5. The enrichment study (148 actives, 5700 decoys) has shown excellent classification ability of our models (AUC > 0.90) for all cases). Conclusion: We found that the kPLS method, in combination with binary fingerprints, is useful for the affinity prediction and the Bcl-2 inhibitors classification. The obtained promising results, methods, and applications highlighted in this study will help us to design more selective Bcl-2 inhibitors with better structural characteristics and improved anti-cancer activity.

Modeling Binary Fingerprint Descriptors With the Superposing Significant Interaction Rules (SSIR) Method

Recently, the superposing significant interaction rules (SSIR) method has been applied in several fields of QSPR to model and establish molecular rankings that correlate dichotomous properties. The origin of the method is in the field of combinatorial chemistry, but it has been shown that the procedure is fast, versatile, and that it can be applied in many other fields. In particular, an example is phospholipidosis modeling taking, as primary descriptors, the binary fingerprints of the molecules. This is the first time SSIR is used to treat this kind of descriptors. The performance achieved is similar to other results found in the literature and, in particular, to the results obtained by authors who considered the same molecular set and descriptors. One of the main advantages of SSIR is that the method acts as an automated variable selector. This allows it to be used almost immediately without prior selection of variables.

Structural Insights into the Molecular Design of HER2 Inhibitors

Background: The present study was aimed at designing some potential candidates as HER2 inhibitors used in breast cancer. Methods: An energy optimized pharmacophore (E-pharmacophore) model was developed and used to screen the molecular databases (such as ASINEX and NCI databases) against a six site (ADHRRR) hypothesis. The shape similarity of the retrieved hits was calculated and then filtered applying ADME and Lipinski’s filters. Further, these hits were docked into the crystal structure of HER2 protein (3W32) using Glide XP protocol to obtain the docking poses and XP gscores. The performance of the virtual screening (VS) methods was evaluated using Schrödinger’s decoy set of 1000 molecules. Ranking of the actives in the VS protocol was assessed by a variety of well-established methods including the average rank of actives, EF, ROC, BEDROC, AUAC, and the RIE. The retrieved hits were submitted to Canvas for generating binary fingerprints (dendritic) to identify structural diversity among the hits and clustered on the basis of Tanimoto coefficient using hierarchical clustering. Results: Seven structurally diverse clusters were selected applying above protocol, having XP gscores >-10, and fitness scores > 1, considering top scoring cluster representative from each cluster. The best scoring hit 355682-ASINEX was submitted to Combiglide to discover some better candidates with improved scores. Finally, structural interaction fingerprint (SIFT) analysis was employed to study the binding interaction, which showed H-bond interaction with Met793, Gln791 and Thr854 residues of HER2 protein. Conclusion: The applied methodology and the retrieved hits could be useful in the design of potent inhibitors of HER2 proteins, commonly found to be expressed in the breast cancer patients.

Interactions Between Weighting Scheme and Similarity Coefficient in Similarity-Based Virtual Screening

Similarity searching is one of the most common methods for ligand-based virtual screening, and is normally carried out using the Tanimoto coefficient with binary fingerprints. However, a recent study has suggested that it may be less appropriate for use with weighted fingerprints in some circumstances. This paper compares the Tanimoto coefficient with other coefficients, and demonstrates that one of these, the cosine coefficient, exhibits a much greater degree of robustness in the face of variations in the nature of the fragment weighting scheme that is being used.