Comprehensive machine learning based study of the chemical space of herbicides

Systematic approaches to machine learning models for predicting pesticide toxicity

Pesticides play an important role in modern agriculture by protecting crops from pests and diseases. However, the negative consequences of pesticides, such as environmental contamination and adverse effects on human and ecological health, underscore the importance of accurate toxicity predictions. To address this issue, artificial intelligence models have emerged as valuable methods for predicting the toxicity of organic compounds. In this review article, we explore the application of machine learning (ML) for pesticide toxicity prediction. This review provides a detailed summary of recent developments, prediction models, and datasets used for pesticide toxicity prediction. In this analysis, we compared the results of several algorithms that predict the harmfulness of various classes of pesticides. Furthermore, this review article identified emerging trends and areas for future direction, showcasing the transformative potential of machine learning in promoting safer pesticide usage and sustainable agriculture.

Molecular Targets of Herbicides and Fungicides─Are There Useful Overlaps for Fungicide Discovery?

New fungicide modes of action are needed for fungicide resistance management strategies. Several commercial herbicide targets found in fungi that are not utilized by commercial fungicides are discussed as possible fungicide molecular targets. These are acetyl CoA carboxylase, acetolactate synthase, 5-enolpyruvylshikimate-3-phosphate synthase, glutamine synthase, phytoene desaturase, protoporphyrinogen oxidase, long-chain fatty acid synthase, dihydropteroate synthase, hydroxyphenyl pyruvate dioxygenase, and Ser/Thr protein phosphatase. Some of the inhibitors of these herbicide targets appear to be either good fungicides or good leads for new fungicides. For example, some acetolactate synthase and dihydropteroate inhibitors are excellent fungicides. There is evidence that some herbicides have indirect benefits to certain crops due to their effects on fungal crop pathogens. Using a pesticide with both herbicide and fungicide activities based on the same molecular target could reduce the total amount of pesticide used. The limitations of such a product are discussed.

Selected herbicides screened for toxicity and analysed as inhibitors of both cholinesterases

Sets of 346 herbicides in use and 163 outdated no longer in use were collected from open access online sources and compared in silico with cholinesterases inhibitors (ChI) and drugs in terms of physicochemical profile and estimated toxic effects on human health. The screening revealed at least one potential adverse consequence for each herbicide class assigned according to their mode of action on weeds. The classes with most toxic warnings were K1, K3/N, F1 and E. The selection of 11 commercial herbicides for in vitro biological tests on human acetylcholinesterase (AChE) and butyrylcholinesterase (BChE), the enzymes involved in neurotoxicity and detoxification of various xenobiotics, respectively, was based mainly on the structural similarity with inhibitors of cholinesterases. Organophosphate anilofos and oxyacetanilide flufenacet were the most potent inhibitors of AChE (25 μM) and BChE (6.4 μM), respectively. Glyphosate, oxadiazon, tembotrione and terbuthylazine were poor inhibitors with an estimated IC₅₀ above 100 μM, while for glyphosate the IC₅₀ was above 1 mM. Generally, all of the selected herbicides inhibited with a slight preference towards BChE. Cytotoxicity assays showed that anilofos, bensulide, butamifos, piperophos and oxadiazon were cytotoxic for hepatocytes (HepG2) and neuroblastoma cell line (SH-SY5Y). Time-independent cytotoxicity accompanied with induction of reactive oxygen species indicated rapid cell death in few hours. Our results based on in silico and in vitro analyses give insight into the potential toxic outcome of herbicides in use and can be applied in the design of new molecules with a less hazardous impact on humans and the environment.

Molecular targets of insecticides and herbicides - Are there useful overlaps?

New insecticide modes of action are needed for insecticide resistance management strategies. The number of molecular targets of commercial herbicides and insecticides are fewer than 35 for both. Few commercial insecticide targets are found in plants, but ten targets of commercial herbicides are found in insects. For several of these commonly held targets, some compounds kill both plants and insects. For example, herbicidal inhibitors of p-hydroxyphenylpyruvate dioxygenase are effective insecticides on blood-fed insects. The glutamine synthetase-inhibiting herbicide glufosinate is insecticidal by the same mechanism of action, inhibition of glutamine synthetase. These and other examples of shared activities of commercial herbicides with insecticides through the same target site are discussed. Compounds with novel herbicide targets shared by insects that are not commercialized as pesticides (such as statins) are also discussed. Compounds that are both herbicidal and insecticidal can be used for insect pests not associated with crops or with crops made resistant to the compounds.

Modern Approaches for the Development of New Herbicides Based on Natural Compounds

Weeds are a permanent component of anthropogenic ecosystems. They require strict control to avoid the accumulation of their long-lasting seeds in the soil. With high crop infestation, many elements of crop production technologies (fertilization, productive varieties, growth stimulators, etc.) turn out to be practically meaningless due to high yield losses. Intensive use of chemical herbicides (CHs) has led to undesirable consequences: contamination of soil and wastewater, accumulation of their residues in the crop, and the emergence of CH-resistant populations of weeds. In this regard, the development of environmentally friendly CHs with new mechanisms of action is relevant. The natural phytotoxins of plant or microbial origin may be explored directly in herbicidal formulations (biorational CHs) or indirectly as scaffolds for nature-derived CHs. This review considers (1) the main current trends in the development of CHs that may be important for the enhancement of biorational herbicides; (2) the advances in the development and practical application of natural compounds for weed control; (3) the use of phytotoxins as prototypes of synthetic herbicides. Some modern approaches, such as computational methods of virtual screening and design of herbicidal molecules, development of modern formulations, and determination of molecular targets, are stressed as crucial to make the exploration of natural compounds more effective.

Drug Chemical Space as a Guide for New Herbicide Development: A Cheminformatic Analysis

Herbicides are critical resources for meeting agricultural demand. While similar in structure and function to pharmaceuticals, the development of new herbicidal mechanisms of action and new scaffolds against known mechanisms of action has been much slower than in pharmaceutical sciences. We hypothesized that this may be due in part to a relative undersampling of possible herbicidal chemistries and set out to test whether this difference in sampling existed and whether increasing the diversity of possible herbicidal chemistries would be likely to result in more efficacious herbicides. To conduct this work, we first identified databases of commercially available herbicides and clinically approved pharmaceuticals. Using these databases, we created a two-dimensional embedding of the chemical, which provides a qualitative visualization of the degree to which each chemotype is distributed within the combined chemical space and shows a moderate degree of overlap between the two sets. Next, we trained several machine learning models to classify herbicides versus drugs based on physicochemical characteristics. The most accurate of these models has an accuracy of 93% with the key differentiating characteristics being the number of polar hydrogens, number of amide bonds, LogP, and polar surface area. We then used several types of scaffold decomposition to quantitatively evaluate the chemical diversity of each molecular family and showed herbicides to have considerably fewer unique structural fragments. Finally, we used molecular docking as an in silico evaluation of further structural diversification in herbicide development. To this end, we identified herbicides with well-characterized binding sites and modified those scaffolds based on similar structural subunits from the drug dataset not present in any commercial herbicide while using the machine-learned model to ensure that required herbicide properties were maintained. Redocking the original and modified scaffolds of several herbicides showed that even this simple design strategy is capable of yielding new molecules with higher predicted affinity for the target enzymes. Overall, we show that herbicides are distinct from drugs based on physicochemical properties but less diverse in their chemistry in a way not governed by these properties. We also demonstrate in silico that increasing the diversity of herbicide scaffolds has the potential to increase potency, potentially reducing the amount needed in agricultural practice.

Anacardic Acid Derivatives Affect the in Vitro Reactions of Photosynthesis

The dichloromethane extract of the cashew nuts from Anacardium occidentale was fractionated by rotation locular countercurrent chromatography aimed at discovering metabolites that could be useful as new models for photosynthesis inhibitors. The chemical fractionation afforded a complex mixture of anacardic acids, which upon catalytic hydrogenation yielded anacardic acid (1). Methylation of 1 via reaction with diazomethane afforded an ester 2. Both compounds were evaluated using polarographic approaches and fluorescence studies of chlorophyll a (ChL a). The in vitro assays informed the decision for the classification of 1 and 2 as Hill reaction inhibitors. Besides that, 1 inhibited the donor side of the PSII, while 2 acted as an energy transfer inhibitor. Therefore, this study is important for the development of herbicides.

Methylglyoxal detoxification pathway - Explored first time for imazethapyr tolerance in lentil (Lens culinaris L.)

Lentil is an important pulses crop but it's short stature and slow growth rate make it vulnerable to weed competition, limiting crop productivity. There is need to identify herbicide tolerant genotypes and their tolerance mechanism. The present investigation was conducted to understand the effect of imazethapyr (IM) treatment on accumulation of methylglyoxal (MG) and its detoxification mechanism in IM-tolerant (LL1397 and LL1612) susceptible (FLIP2004-7L and PL07) genotypes sown under control (weed free), weedy check (weeds were growing with crop) and sprayed with imazethapyr. The enzymes of glyoxalase pathway (glyoxalase I, II and III) and non glyoxalase pathway (methylglyoxal reductase), lactate dehydrogenase (LDH), glutathione content, gamma-glutamyl-cysteine synthetase (γ-GCS) were estimated in lentil genotypes at different days after spray. Higher activities of glyoxalase I, II and III and MGR along with the increased glutathione content (GSH) content in LL1397 and LL1612 under IM treatment as compared to FLIP2004-7L and PL07 might be responsible for lowering MG accumulation and increasing lactate content, which is end product of these pathways. Enhanced LDH activity in LL1397 and LL1612 might be responsible for energy production via TCA cycle that might be responsible for growth and recovery of tolerant genotypes after IM treatment. Higher γ-GCS activity in tolerant genotypes led to increased glutathione content required for glyoxalase pathway. However, decreased activities of glyoxalase enzymes and MGR in susceptible genotypes result in MG accumulation which limit plant growth. This is the first ever study elucidating the role of MG detoxification pathway conferring IM tolerance in lentil.

Biological Membrane-Penetrating Peptides: Computational Prediction and Applications

Peptides comprise a versatile class of biomolecules that present a unique chemical space with diverse physicochemical and structural properties. Some classes of peptides are able to naturally cross the biological membranes, such as cell membrane and blood-brain barrier (BBB). Cell-penetrating peptides (CPPs) and blood-brain barrier-penetrating peptides (B3PPs) have been explored by the biotechnological and pharmaceutical industries to develop new therapeutic molecules and carrier systems. The computational prediction of peptides’ penetration into biological membranes has been emerged as an interesting strategy due to their high throughput and low-cost screening of large chemical libraries. Structure- and sequence-based information of peptides, as well as atomistic biophysical models, have been explored in computer-assisted discovery strategies to classify and identify new structures with pharmacokinetic properties related to the translocation through biomembranes. Computational strategies to predict the permeability into biomembranes include cheminformatic filters, molecular dynamics simulations, artificial intelligence algorithms, and statistical models, and the choice of the most adequate method depends on the purposes of the computational investigation. Here, we exhibit and discuss some principles and applications of these computational methods widely used to predict the permeability of peptides into biomembranes, exhibiting some of their pharmaceutical and biotechnological applications.

Hierarchical cluster analysis of herbicide modes of action reveals distinct classes of multiple resistance in weeds

BACKGROUND

The number of weed species resistant to multiple herbicide modes of action (MoAs) has increased over the last 30 years and may in the future render existing herbicide MoAs obsolete for many cropping systems. Yet few predictive tools exist to manage this risk. Using a worldwide dataset of weed species resistant to multiple herbicide MoAs, hierarchical clustering was used to classify MoAs into similar groups in relation to the suite of resistant weed species they have in common. Network analyses then were used to explore the relative importance of species prevalence and similarity in cluster patterns.

RESULTS

Hierarchical clustering identified three similarly sized clusters of herbicide MoAs that were linked by the co-occurrence of resistant weeds: Herbicide Resistance Action Committee (HRAC) groups 2, 4, 5 and 9; HRAC groups 12, 14 and 15; and HRAC groups 1, 3 and 22. Cluster membership was consistent with similarities in the physiological or biochemical target of the herbicide MoAs. Network analyses revealed that the number of weed species resistant to two different MoAs was related to the number of weeds known to be resistant to each individual herbicide MoA.

CONCLUSIONS

Hierarchical cluster analysis provided new insights into the risk of weeds becoming resistant to more than one herbicide MoA. By clustering herbicide MoAs into three distinct groups, the potential exists for farmers to manage resistance by rotating herbicides between rather than within clusters, as far as crop, weed and environmental conditions allow.

cropCSM: designing safe and potent herbicides with graph-based signatures

Herbicides have revolutionised weed management, increased crop yields and improved profitability allowing for an increase in worldwide food security. Their widespread use, however, has also led to a rise in resistance and concerns about their environmental impact. Despite the need for potent and safe herbicidal molecules, no herbicide with a new mode of action has reached the market in 30 years. Although development of computational approaches has proven invaluable to guide rational drug discovery pipelines, leading to higher hit rates and lower attrition due to poor toxicity, little has been done in contrast for herbicide design. To fill this gap, we have developed cropCSM, a computational platform to help identify new, potent, nontoxic and environmentally safe herbicides. By using a knowledge-based approach, we identified physicochemical properties and substructures enriched in safe herbicides. By representing the small molecules as a graph, we leveraged these insights to guide the development of predictive models trained and tested on the largest collected data set of molecules with experimentally characterised herbicidal profiles to date (over 4500 compounds). In addition, we developed six new environmental and human toxicity predictors, spanning five different species to assist in molecule prioritisation. cropCSM was able to correctly identify 97% of herbicides currently available commercially, while predicting toxicity profiles with accuracies of up to 92%. We believe cropCSM will be an essential tool for the enrichment of screening libraries and to guide the development of potent and safe herbicides. We have made the method freely available through a user-friendly webserver at http://biosig.unimelb.edu.au/crop_csm.