Machine Learning for Molecular Modelling in Drug Design

Rationale: Low speed and flexibility of most LC-MS/MS approaches in early drug discovery delays sample analysis from routine in vivo studies within the same day of measurements. A highthroughput platform for the rapid quantification of drug compounds in various in vivo assays was developed and established in routine bioanalysis. Methods: Automated selection of an efficient and adequate LC method was realized by autonomous sample qualification for ultrafast batch gradients (9 s/sample) or for fast linear gradients (45 s/sample) if samples require chromatography. The hardware and software components of our Rapid and Integrated Analysis System (RIAS) were streamlined for increased analytical throughput via state-of-the-art automation while keeping high analytical quality. Results: Online decision-making was based on a quick assay suitability test (AST) based on a small and dedicated sample set evaluated by two different strategies. 84% of the acquired data points were within ±30% accuracy and 93% of the deviations between the lower limit of quantitation (LLOQ) values were ≤2-fold compared to standard LC-MS/MS systems while speed, flexibility and overall automation was significantly improved. Conclusions: The developed platform provided an analysis time of only 10 min (batch-mode) and 50 min (gradient-mode) per standard pharmacokinetic (PK) study (62 injections). Automation, data evaluation and results handling were optimized to pave the way for machine learning based decision-making regarding the evaluation strategy of the AST

Download Full-text

Ligand- and Structure-Based Drug Design and Optimization using KNIME

Current Medicinal Chemistry ◽

10.2174/0929867326666190409141016 ◽

2020 ◽

Vol 27 (38) ◽

pp. 6458-6479 ◽

Cited By ~ 4

Author(s):

Michael P. Mazanetz ◽

Charlotte H.F. Goode ◽

Ewa I. Chudyk

Keyword(s):

Drug Discovery ◽

Drug Design ◽

Candidate Selection ◽

Structure Based Drug Design ◽

Design And Optimization ◽

Open Source Data ◽

Source Data ◽

Recent Developments ◽

Hit Identification ◽

Early Drug

In recent years there has been a paradigm shift in how data is being used to progress early drug discovery campaigns from hit identification to candidate selection. Significant developments in data mining methods and the accessibility of tools for research scientists have been instrumental in reducing drug discovery timelines and in increasing the likelihood of a chemical entity achieving drug development milestones. KNIME, the Konstanz Information Miner, is a leading open source data analytics platform and has supported drug discovery endeavours for over a decade. KNIME provides a rich palette of tools supported by an extensive community of contributors to enable ligandand structure-based drug design. This review will examine recent developments within the KNIME platform to support small-molecule drug design and provide a perspective on the challenges and future developments within this field.

Download Full-text

Faculty Opinions recommendation of Model-based Target Pharmacology Assessment (mTPA): An Approach Using PBPK/PD Modeling and Machine Learning to Design Medicinal Chemistry and DMPK Strategies in Early Drug Discovery.

Faculty Opinions – Post-Publication Peer Review of the Biomedical Literature ◽

10.3410/f.739749744.793584168 ◽

2021 ◽

Author(s):

John Lowe

Keyword(s):

Machine Learning ◽

Drug Discovery ◽

Medicinal Chemistry ◽

Model Based ◽

Early Drug

Download Full-text

A Structure-Based Drug Discovery Paradigm

International Journal of Molecular Sciences ◽

10.3390/ijms20112783 ◽

2019 ◽

Vol 20 (11) ◽

pp. 2783 ◽

Cited By ~ 50

Author(s):

Maria Batool ◽

Bilal Ahmad ◽

Sangdun Choi

Keyword(s):

Machine Learning ◽

Deep Learning ◽

Drug Discovery ◽

Drug Design ◽

De Novo ◽

Learning Tools ◽

Statistical Machine Learning ◽

Lead Discovery ◽

De Novo Drug Design ◽

Structure Based Drug Design

Structure-based drug design is becoming an essential tool for faster and more cost-efficient lead discovery relative to the traditional method. Genomic, proteomic, and structural studies have provided hundreds of new targets and opportunities for future drug discovery. This situation poses a major problem: the necessity to handle the “big data” generated by combinatorial chemistry. Artificial intelligence (AI) and deep learning play a pivotal role in the analysis and systemization of larger data sets by statistical machine learning methods. Advanced AI-based sophisticated machine learning tools have a significant impact on the drug discovery process including medicinal chemistry. In this review, we focus on the currently available methods and algorithms for structure-based drug design including virtual screening and de novo drug design, with a special emphasis on AI- and deep-learning-based methods used for drug discovery.

Download Full-text

Nonadditivity in Public and Inhouse Data – Implications for Drug Design

10.21203/rs.3.rs-127217/v1 ◽

2020 ◽

Author(s):

Dea Gogishvili ◽

Eva Nittinger ◽

Christian Margreitter ◽

Christian Tyrchan

Keyword(s):

Machine Learning ◽

Drug Discovery ◽

Drug Design ◽

Measurement Errors ◽

Rational Drug Design ◽

Data Sets ◽

Structure Activity ◽

Rational Drug ◽

Public Data ◽

Sar Analysis

Abstract Numerous ligand-based drug discovery projects are based on structure-activity relationship (SAR) analysis, such as Free-Wilson (FW) or matched molecular pair (MMP) analysis. Intrinsically they assume linearity and additivity of substituent contributions. These techniques are challenged by nonadditivity (NA) in protein-ligand binding where the change of two functional groups in one molecule results in much higher or lower activity than expected from the respective single changes. Identifying nonlinear cases and possible underlying explanations is crucial for a drug design project since it might influence which lead to follow. By systematically analyzing all AstraZeneca (AZ) inhouse compound data and publicly available ChEMBL25 bioactivity data, we show significant NA events in almost every second assay among the inhouse and once in every third assay in public data sets. Furthermore, 9.4% of all compounds of the AZ database and 5.1% from public sources display significant additivity shifts indicating important SAR features or fundamental measurement errors. Using NA data in combination with machine learning showed that nonadditive data is challenging to predict and even the addition of nonadditive data into training did not result in an increase in predictivity. Overall, NA analysis should be applied on a regular basis in many areas of computational chemistry and can further improve rational drug design.

Download Full-text

Artificial intelligence in drug design: algorithms, applications, challenges and ethics

Future Drug Discovery ◽

10.4155/fdd-2020-0028 ◽

2021 ◽

pp. FDD59

Author(s):

Alya A Arabi

Keyword(s):

Artificial Intelligence ◽

Machine Learning ◽

Drug Discovery ◽

Drug Design ◽

Chemical Properties ◽

Binding Affinities ◽

Future Perspectives ◽

Ethical Considerations ◽

Brain Barrier Permeability ◽

And Chemical Properties

The discovery paradigm of drugs is rapidly growing due to advances in machine learning (ML) and artificial intelligence (AI). This review covers myriad faces of AI and ML in drug design. There is a plethora of AI algorithms, the most common of which are summarized in this review. In addition, AI is fraught with challenges that are highlighted along with plausible solutions to them. Examples are provided to illustrate the use of AI and ML in drug discovery and in predicting drug properties such as binding affinities and interactions, solubility, toxicology, blood–brain barrier permeability and chemical properties. The review also includes examples depicting the implementation of AI and ML in tackling intractable diseases such as COVID-19, cancer and Alzheimer’s disease. Ethical considerations and future perspectives of AI are also covered in this review.

Download Full-text

An Ultrafast and Flexible LC-MS/MS System Paves the Way for Machine Learning Driven Sample Processing and Data Evaluation in Early Drug Discovery

10.26434/chemrxiv.13917179 ◽

2021 ◽

Author(s):

Tim Häbe ◽

Christian Späth ◽

Steffen Schrade ◽

Wolfgang Jörg ◽

Roderich Süssmuth ◽

...

Keyword(s):

Machine Learning ◽

Decision Making ◽

Drug Discovery ◽

Data Evaluation ◽

In Vivo Studies ◽

Integrated Analysis ◽

Limit Of Quantitation ◽

Early Drug ◽

The Way

Rationale: Low speed and flexibility of most LC-MS/MS approaches in early drug discovery delays sample analysis from routine in vivo studies within the same day of measurements. A highthroughput platform for the rapid quantification of drug compounds in various in vivo assays was developed and established in routine bioanalysis. Methods: Automated selection of an efficient and adequate LC method was realized by autonomous sample qualification for ultrafast batch gradients (9 s/sample) or for fast linear gradients (45 s/sample) if samples require chromatography. The hardware and software components of our Rapid and Integrated Analysis System (RIAS) were streamlined for increased analytical throughput via state-of-the-art automation while keeping high analytical quality. Results: Online decision-making was based on a quick assay suitability test (AST) based on a small and dedicated sample set evaluated by two different strategies. 84% of the acquired data points were within ±30% accuracy and 93% of the deviations between the lower limit of quantitation (LLOQ) values were ≤2-fold compared to standard LC-MS/MS systems while speed, flexibility and overall automation was significantly improved. Conclusions: The developed platform provided an analysis time of only 10 min (batch-mode) and 50 min (gradient-mode) per standard pharmacokinetic (PK) study (62 injections). Automation, data evaluation and results handling were optimized to pave the way for machine learning based decision-making regarding the evaluation strategy of the AST

Download Full-text