skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.


Title: Classifying the toxicity of pesticides to honey bees via support vector machines with random walk graph kernels
Pesticides benefit agriculture by increasing crop yield, quality, and security. However, pesticides may inadvertently harm bees, which are valuable as pollinators. Thus, candidate pesticides in development pipelines must be assessed for toxicity to bees. Leveraging a dataset of 382 molecules with toxicity labels from honey bee exposure experiments, we train a support vector machine (SVM) to predict the toxicity of pesticides to honey bees. We compare two representations of the pesticide molecules: (i) a random walk feature vector listing counts of length- L walks on the molecular graph with each vertex- and edge-label sequence and (ii) the Molecular ACCess System (MACCS) structural key fingerprint (FP), a bit vector indicating the presence/absence of a list of pre-defined subgraph patterns in the molecular graph. We explicitly construct the MACCS FPs but rely on the fixed-length- L random walk graph kernel (RWGK) in place of the dot product for the random walk representation. The L-RWGK-SVM achieves an accuracy, precision, recall, and F1 score (mean over 2000 runs) of 0.81, 0.68, 0.71, and 0.69, respectively, on the test data set—with L = 4 being the mode optimal walk length. The MACCS-FP-SVM performs on par/marginally better than the L-RWGK-SVM, lends more interpretability, but varies more in performance. We interpret the MACCS-FP-SVM by illuminating which subgraph patterns in the molecules tend to strongly push them toward the toxic/non-toxic side of the separating hyperplane.  more » « less
Award ID(s):
1920945
PAR ID:
10351745
Author(s) / Creator(s):
; ; ;
Date Published:
Journal Name:
The Journal of Chemical Physics
Volume:
157
Issue:
3
ISSN:
0021-9606
Page Range / eLocation ID:
034102
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract Facing the continuous emergence of new psychoactive substances (NPS) and their threat to public health, more effective methods for NPS prediction and identification are critical. In this study, the pharmacological affinity fingerprints (Ph-fp) of NPS compounds were predicted by Random Forest classification models using bioactivity data from the ChEMBL database. The binaryPh-fpis the vector consisting of a compound’s activity against a list of molecular targets reported to be responsible for the pharmacological effects of NPS. Their performance in similarity searching and unsupervised clustering was assessed and compared to 2D structure fingerprints Morgan and MACCS (1024-bits ECFP4 and 166-bits SMARTS-based MACCS implementation of RDKit). The performance in retrieving compounds according to their pharmacological categorizations is influenced by the predicted active assay counts inPh-fpand the choice of similarity metric. Overall, the comparative unsupervised clustering analysis suggests the use of a classification model with Morgan fingerprints as input for the construction ofPh-fp. This combination gives satisfactory clustering performance based on external and internal clustering validation indices. 
    more » « less
  2. Insecticides are a major tool for controlling pest species. Their widespread use results in damage to non-targeted insects, with honey bees particularly at risk. During foraging, honey bees learn and remember floral charac teristics that are associated with food. As insect pollinators, honey bees inadvertently contact chemicals which can have multiple negative impacts. The toxicity of two insecticides from different classes, ethion (47.79 mg a.i. L − 1 ) and hexaflumuron (500 mg a.i.L − 1 ), on learning, memory, and sensory perception were evaluated. We found that oral exposure to ethion had adverse effects on learned proboscis extension toward reward-associated odors and colors. In addition, we showed reduced sucrose consumption and sucrose responsiveness after expo sure. Hexaflumuron also impaired olfactory learning and memory and decreased responsiveness to sucrose and water. Exposure to sub-lethal concentration of the cholinergic organophosphate insecticide, ethion (47.79 mg a.i. L − 1 ), and the field-recommended concentration of hexaflumuron (500 mg a.i.L − 1 ), significantly impaired behavior involved in foraging. Our results suggest that several behavioral characteristics of honey bees be evaluated when testing an insecticide rather than relying on just one behavioral measure. 
    more » « less
  3. Abstract Pesticides and parasites have each been linked to increased mortality in western honey bees (Apis mellifera). Currently, it is uncertain if one makes the other worse; several studies have tested for potential synergistic stressor effects, but results have been mixed.Here, we use a hierarchical meta‐analysis of 63 experiments from 26 studies to gain a clearer view of the combined effects of parasites and pesticides on honey bee health.We found that combined pesticide–parasite treatments do tend to be deadlier than uncombined treatments but are significantly less deadly than predicted additive or multiplicative effects. In other words, combined treatment effects are not synergistic, but antagonistic.Much of the previous uncertainty about the combined effects of pesticides and parasites on honey bee health can be attributed to a bias in the previous research against stressor antagonism; many researchers have excluded the possibility of antagonism a priori.Synthesis and applications. Meta‐analysis shows that when honey bees are stressed by a combination of pesticides and parasites, the combined stress effect is antagonistic, that is, less than the sum of its parts. A better understanding of the mechanisms underlying this antagonism could prove critical for effective management of honey bee health. 
    more » « less
  4. Brunet, Johanne (Ed.)
    Abstract Honey bees (Apis mellifera L. Hymeoptera: Apidae) use hydrogen peroxide (synthesized by excreted glucose oxidase) as an important component of social immunity. However, both tolerance of hydrogen peroxide and the production of glucose oxidase in honey is costly. Hydrogen peroxide may also be encountered by honey bees at high concentrations in nectar while foraging, however despite its presence both in their foraged and stored foods, it is unclear if and how bees monitor concentrations of, and their behavioral responses to, hydrogen peroxide. The costs of glucose oxidase production and the presence of hydrogen peroxide in both nectar and honey suggest hypotheses that honey bees preferentially forage on hydrogen peroxide supplemented feed syrups at certain concentrations, and avoid feed syrups supplemented with hydrogen peroxide at concentrations above some tolerance threshold. We test these hypotheses and find that, counter to expectation, honey bees avoid glucose solutions supplemented with field-relevant hydrogen peroxide concentrations and either avoid or don’t differentiate supplemented sucrose solutions when given choice assays. This is despite honey bees showing high tolerance for hydrogen peroxide in feed solutions, with no elevated mortality until concentrations of hydrogen peroxide exceed 1% (v/v) in solution, with survival apparent even at concentrations up to 10%. The behavioral interaction of honey bees with hydrogen peroxide during both within-colony synthesis in honey and when foraging on nectar therefore likely relies on interactions with other indicator molecules, and maybe constrained evolutionarily in its plasticity, representing a constitutive immune mechanism. 
    more » « less
  5. Abstract Honey bees are critical pollinators in ecosystems and agriculture, but their numbers have significantly declined. Declines in pollinator populations are thought to be due to multiple factors including habitat loss, climate change, increased vulnerability to disease and parasites, and pesticide use. Neonicotinoid pesticides are agonists of insect nicotinic cholinergic receptors, and sub-lethal exposures are linked to reduced honey bee hive survival. Honey bees are highly dependent on circadian clocks to regulate critical behaviors, such as foraging orientation and navigation, time-memory for food sources, sleep, and learning/memory processes. Because circadian clock neurons in insects receive light input through cholinergic signaling we tested for effects of neonicotinoids on honey bee circadian rhythms and sleep. Neonicotinoid ingestion by feeding over several days results in neonicotinoid accumulation in the bee brain, disrupts circadian rhythmicity in many individual bees, shifts the timing of behavioral circadian rhythms in bees that remain rhythmic, and impairs sleep. Neonicotinoids and light input act synergistically to disrupt bee circadian behavior, and neonicotinoids directly stimulate wake-promoting clock neurons in the fruit fly brain. Neonicotinoids disrupt honey bee circadian rhythms and sleep, likely by aberrant stimulation of clock neurons, to potentially impair honey bee navigation, time-memory, and social communication. 
    more » « less