Drug discovery using very large numbers of patents. General strategy with extensive use of match and edit operations

Glass Box and Black Box Machine Learning Approaches to Exploit Compositional Descriptors of Molecules in Drug Discovery and Aid the Medicinal Chemist

The synthetic medicinal chemist plays a vital role in drug discovery. Today there are AI tools to guide next syntheses, but many are "Black Boxes" (BB). One learns little more than the prediction made. There are now also AI methods emphasizing visibility and "explainability" (thus explainable AI or XAI) that could help when "compositional data" are used, but they often still start from seemingly arbitrary learned weights and lack familiar probabilistic measures based on observation and counting from the outset. If probabilistic methods were used in a complementary way with BB methods and demonstrated comparable predictive power, they would provide guidelines about what groups to include and avoid in next syntheses and quantify the relationships in probabilistic terms. These points are demonstrated by blind test comparison of two main types of BB methods and a probabilistic "Glass Box" (GB) method new outside of medicine, but which appears well suited to the above. Because many probabilities can be involved, emphasis is on the predictive power of its simplest explanatory models. There are usually more inactive compounds by orders of magnitude, often a problem for machine learning methods. However, the approaches used here appear to work well for such "real world data".

ReBADD-SE: Multi-objective molecular optimisation using SELFIES fragment and off-policy self-critical sequence training

The discovery of drugs to selectively remove disease-related cells is challenging in computer-aided drug design. Many studies have proposed multi-objective molecular generation methods and demonstrated their superiority using the public benchmark dataset for kinase inhibitor generation tasks. However, the dataset does not contain many molecules that violate Lipinski's rule of five. Thus, it remains unclear whether existing methods are effective in generating molecules violating the rule, such as navitoclax. To address this, we analysed the limitations of existing methods and propose a multi-objective molecular generation method with a novel parsing algorithm for molecular string representation and a modified reinforcement learning method for the efficient training of multi-objective molecular optimisation. The proposed model had success rates of 84% in GSK3b+JNK3 inhibitor generation and 99% in Bcl-2 family inhibitor generation tasks.

The New Coronavirus (SARS-CoV-2): A Comprehensive Review on Immunity and the Application of Bioinformatics and Molecular Modeling to the Discovery of Potential Anti-SARS-CoV-2 Agents

On March 11, 2020, the World Health Organization (WHO) officially declared the outbreak caused by the new coronavirus (SARS-CoV-2) a pandemic. The rapid spread of the disease surprised the scientific and medical community. Based on the latest reports, news, and scientific articles published, there is no doubt that the coronavirus has overloaded health systems globally. Practical actions against the recent emergence and rapid expansion of the SARS-CoV-2 require the development and use of tools for discovering new molecular anti-SARS-CoV-2 targets. Thus, this review presents bioinformatics and molecular modeling strategies that aim to assist in the discovery of potential anti-SARS-CoV-2 agents. Besides, we reviewed the relationship between SARS-CoV-2 and innate immunity, since understanding the structures involved in this infection can contribute to the development of new therapeutic targets. Bioinformatics is a technology that assists researchers in coping with diseases by investigating genetic sequencing and seeking structural models of potential molecular targets present in SARS-CoV2. The details provided in this review provide future points of consideration in the field of virology and medical sciences that will contribute to clarifying potential therapeutic targets for anti-SARS-CoV-2 and for understanding the molecular mechanisms responsible for the pathogenesis and virulence of SARS-CoV-2.

COVID-19 Coronavirus spike protein analysis for synthetic vaccines, a peptidomimetic antagonist, and therapeutic drugs, and analysis of a proposed achilles' heel conserved region to minimize probability of escape mutations and drug resistance

This paper continues a recent study of the spike protein sequence of the COVID-19 virus (SARS-CoV-2). It is also in part an introductory review to relevant computational techniques for tackling viral threats, using COVID-19 as an example. Q-UEL tools for facilitating access to knowledge and bioinformatics tools were again used for efficiency, but the focus in this paper is even more on the virus. Subsequence KRSFIEDLLFNKV of the S2′ spike glycoprotein proteolytic cleavage site continues to appear important. Here it is shown to be recognizable in the common cold coronaviruses, avian coronaviruses and possibly as traces in the nidoviruses of reptiles and fish. Its function or functions thus seem important to the coronaviruses. It might represent SARS-CoV-2 Achilles’ heel, less likely to acquire resistance by mutation, as has happened in some early SARS vaccine studies discussed in the previous paper. Preliminary conformational analysis of the receptor (ACE2) binding site of the spike protein is carried out suggesting that while it is somewhat conserved, it appears to be more variable than KRSFIEDLLFNKV. However compounds like emodin that inhibit SARS entry, apparently by binding ACE2, might also have functions at several different human protein binding sites. The enzyme 11β-hydroxysteroid dehydrogenase type 1 is again argued to be a convenient model pharmacophore perhaps representing an ensemble of targets, and it is noted that it occurs both in lung and alimentary tract. Perhaps it benefits the virus to block an inflammatory response by inhibiting the dehydrogenase, but a fairly complex web involves several possible targets.

•

This paper “drills down” into the studies of the author's previous COVID-19 paper.

•

Designing vaccine and drugs must seek to avoid escape mutations.

•

Subsequence KRSFIEDLLFNKV seems recognizable across many coronaviruses.

•

The ACE2 binding domain is a target, but shows variation.

•

A steroid dehydrogenase is argued to remain an interesting model pharmacophore.

Computers and viral diseases. Preliminary bioinformatics studies on the design of a synthetic vaccine and a preventative peptidomimetic antagonist against the SARS-CoV-2 (2019-nCoV, COVID-19) coronavirus

This paper concerns study of the genome of the Wuhan Seafood Market isolate believed to represent the causative agent of the disease COVID-19. This is to find a short section or sections of viral protein sequence suitable for preliminary design proposal for a peptide synthetic vaccine and a peptidomimetic therapeutic, and to explore some design possibilities. The project was originally directed towards a use case for the Q-UEL language and its implementation in a knowledge management and automated inference system for medicine called the BioIngine, but focus here remains mostly on the virus itself. However, using Q-UEL systems to access relevant and emerging literature, and to interact with standard publically available bioinformatics tools on the Internet, did help quickly identify sequences of amino acids that are well conserved across many coronaviruses including 2019-nCoV. KRSFIEDLLFNKV was found to be particularly well conserved in this study and corresponds to the region around one of the known cleavage sites of the SARS virus that are believed to be required for virus activation for cell entry. This sequence motif and surrounding variations formed the basis for proposing a specific synthetic vaccine epitope and peptidomimetic agent. The work can, nonetheless, be described in traditional bioinformatics terms, and readily reproduced by others, albeit with the caveat that new data and research into 2019-nCoV is emerging and evolving at an explosive pace. Preliminary studies using molecular modeling and docking, and in that context the potential value of certain known herbal extracts, are also described.

•

Bioinformatics studies are carried out on the COVID-19 virus.

•

A sequence motif KRSFIEDLLFNKV is of particular interest.

•

Based on the above, synthetic peptides are designed.

•

Preliminary considerations are also given to non-peptide organic molecules.

Extension of the Quantum Universal Exchange Language to precision medicine and drug lead discovery. Preliminary example studies using the mitochondrial genome

The Quantum Universal Exchange Language (Q-UEL) based on Dirac notation and algebra from quantum mechanics, along with its associated data mining and Hyperbolic Dirac Net (HDN) for probabilistic inference, has proven to be a useful architectural principle for knowledge management, analysis and prediction systems in medicine. It has been described in several papers; here is described its extension to clinical genomics and precision medicine. Two use cases are studied: (a) bioinformatics in clinical decision support especially for risk for type 2 diabetes using mitochondrial patient DNA sequences, and (b) bioinformatics and computational biology (conformational) research examples related to drug discovery involving the recently discovered class of mitochondrial derived peptides (MDPs). MDPs were surprising when first discovered as coded in small open reading frames (sORFs), and are emerging as having a fundamental role in metabolic control, longevity and disease. This project originally represented a language specification study relating to what information related to genomics is essential or useful to carry, and what processing will be needed. However, novel aspects introduced or discovered include the HDN-like neural nets and their use, along with more established methods, for prediction of type 2 diabetes, and in particular for proposals for over 80 natural MDPs most of which that have not previously been described at the time of the study, as potential drug lead targets. Also, use of many medical records with simulated joining of mtDNA as performance tests led to some insightful observations regarding the behavior of HDN predictions where independent factors are involved.

Studies in the extensively automatic construction of large odds-based inference networks from structured data. Examples from medical, bioinformatics, and health insurance claims data

Theoretical and methodological principles are presented for the construction of very large inference nets for odds calculations, composed of hundreds or many thousands or more of elements, in this paper generated by structured data mining. It is argued that the usual small inference nets can sometimes represent rather simple, arbitrary estimates. Examples of applications in clinical and public health data analysis, medical claims data and detection of irregular entries, and bioinformatics data, are presented. Construction of large nets benefits from application of a theory of expected information for sparse data and the Dirac notation and algebra. The extent to which these are important here is briefly discussed. Purposes of the study include (a) exploration of the properties of large inference nets and a perturbation and tacit conditionality models, (b) using these to propose simpler models including one that a physician could use routinely, analogous to a "risk score", (c) examination of the merit of describing optimal performance in a single measure that combines accuracy, specificity, and sensitivity in place of a ROC curve, and (d) relationship to methods for detecting anomalous and potentially fraudulent data.

Studies in using a universal exchange and inference language for evidence based medicine. Semi-automated learning and reasoning for PICO methodology, systematic review, and environmental epidemiology

The Q-UEL language of XML-like tags and the associated software applications are providing a valuable toolkit for Evidence Based Medicine (EBM). In this paper the already existing applications, data bases, and tags are brought together with new ones. The particular Q-UEL embodiment used here is the BioIngine. The main challenge is one of bringing together the methods of symbolic reasoning and calculative probabilistic inference that underlie EBM and medical decision making. Some space is taken to review this background. The unification is greatly facilitated by Q-UEL's roots in the notation and algebra of Dirac, and by extending Q-UEL into the Wolfram programming environment. Further, the overall problem of integration is also a relatively simple one because of the nature of Q-UEL as a language for interoperability in healthcare and biomedicine, while the notion of workflow is facilitated because of the EBM best practice known as PICO. What remains difficult is achieving a high degree of overall automation because of a well-known difficulty in capturing human expertise in computers: the Feigenbaum bottleneck.

Text mining patents for biomedical knowledge

Biomedical text mining of scientific knowledge bases, such as Medline, has received much attention in recent years. Given that text mining is able to automatically extract biomedical facts that revolve around entities such as genes, proteins, and drugs, from unstructured text sources, it is seen as a major enabler to foster biomedical research and drug discovery. In contrast to the biomedical literature, research into the mining of biomedical patents has not reached the same level of maturity. Here, we review existing work and highlight the associated technical challenges that emerge from automatically extracting facts from patents. We conclude by outlining potential future directions in this domain that could help drive biomedical research and drug discovery.

Hyperbolic Dirac Nets for medical decision support. Theory, methods, and comparison with Bayes Nets

We recently introduced the concept of a Hyperbolic Dirac Net (HDN) for medical inference on the grounds that, while the traditional Bayes Net (BN) is popular in medicine, it is not suited to that domain: there are many interdependencies such that any "node" can be ultimately conditional upon itself. A traditional BN is a directed acyclic graph by definition, while the HDN is a bidirectional general graph closer to a diffuse "field" of influence. Cycles require bidirectionality; the HDN uses a particular type of imaginary number from Dirac׳s quantum mechanics to encode it. Comparison with the BN is made alongside a set of recipes for converting a given BN to an HDN, also adding cycles that do not usually require reiterative methods. This conversion is called the P-method. Conversion to cycles can sometimes be difficult, but more troubling was that the original BN had probabilities needing adjustment to satisfy realism alongside the important property called "coherence". The more general and simpler K-method, not dependent on the BN, is usually (but not necessarily) derived by data mining, and is therefore also introduced. As discussed, BN developments may converge to an HDN-like concept, so it is reasonable to consider the HDN as a BN extension.

Suggestions for a Web based universal exchange and inference language for medicine

Mining biomedical and pharmaceutical data generates huge numbers of interacting probabilistic statements for inference, which can be supported by mining Web text sources. This latter can also be probabilistic, in a sense described in this report. However, the diversity of tools for probabilistic inference is troublesome, suggesting a need for a unifying best practice. Physicists often claim that quantum mechanics is the universal best practice for probabilistic reasoning. We discuss how the Dirac notation and algebra suggest the form and algebraic and semantic meaning of XML-like Web tags for a clinical and biomedical universal exchange language formulated to make sense directly to the eye of the physician and biomedical researcher.