Tracing the origin of honey products based on metagenomics and machine learning

Recent advances of machine learning in the geographical origin traceability of food and agro-products: A review

The geographical origin traceability of food and agro-products has been attracted worldwide. Especially with the rise of machine learning (ML) technology, it provides cutting-edge solutions to erstwhile intractable issues to identify the origin of food and agro-products. By utilizing advanced algorithms, ML can extract feature information of food and agro-products that is closely related to origin and, more accurately, identify and trace their origins, which is of great significance to the entire food and agriculture industry. This paper provides a comprehensive overview of the state-of-the-art applications of ML in the geographical origin traceability of food and agro-products. First, commonly used ML methods are summarized. The paper then outlines the whole process of preparation for modeling, model training as well as model evaluation for building traceability models-based ML. Finally, recent applications of ML combined with different traceability techniques in the field of food and agro-products are revisited. Although ML has made many achievements in solving the geographical origin traceability problem of food and agro-products, it still has great development potential. For example, the application of ML is yet insufficient in the geographical origin traceability using DNA or computer vision techniques. The ability of ML to predict the geographical origin of food and agro-products can be further improved, for example, by increasing model interpretability, incorporating data fusion strategies, and others.

Tracing the genealogy origin of geographic populations based on genomic variation and deep learning

Assigning a query individual animal or plant to its derived population is a prime task in diverse applications related to organismal genealogy. Such endeavors have conventionally relied on short DNA sequences under a phylogenetic framework. These methods naturally show constraints when the inferred population sources are ambiguously phylogenetically structured, a scenario demanding substantially more informative genetic signals. Recent advances in cost-effective production of whole-genome sequences and artificial intelligence have created an unprecedented opportunity to trace the population origin for essentially any given individual, as long as the genome reference data are comprehensive and standardized. Here, we developed a convolutional neural network method to identify population origins using genomic SNPs. Three empirical datasets (an Asian honeybee, a red fire ant, and a chicken datasets) and two simulated populations are used for the proof of concepts. The performance tests indicate that our method can accurately identify the genealogy origin of query individuals, with success rates ranging from  93 % to 100 %. We further showed that the accuracy of the model can be significantly increased by refining the informative sites through FST filtering. Our method is robust to configurations related to batch sizes and epochs, whereas model learning benefits from the setting of a proper preset learning rate. Moreover, we explained the importance score of key sites for algorithm interpretability and credibility, which has been largely ignored. We anticipate that by coupling genomics and deep learning, our method will see broad potential in conservation and management applications that involve natural resources, invasive pests and weeds, and illegal trades of wildlife products.

Tracing the origin and authenticity of Spanish PDO honey using metagenomics and machine learning

The Protected Designation of Origin (PDO) indication for foods intends to guarantee the conditions of production and the geographical origin of regional products within the European Union. Honey products are widely consumed due to their health-promoting properties and there is a general interest in tracing their authenticity. In this regard, metagenomics sequencing and machine learning (ML) have been proposed as complementary technologies to improve the traceability methods of foods. Therefore, the aim of this study was to analyze the metagenomic profiles of Spanish honeys from three different PDOs (Granada, Tenerife and Villuercas-Ibores), and compare them with non-PDO honeys using ML models (PLS, RF, LOGITBOOST, and NNET). According to the results obtained, non-PDO honeys and Granada PDO showed higher beta diversity values than Tenerife and Villuercas-Ibores PDOs. ML classification of honey products allowed the identification of different microbial biomarkers of the geographical origin of honeys: Lactobacillus kunkeei, Parasaccharibacter apium and Lactobacillus helsingborgensis for PDO honeys and Paenibacillus larvae, Lactobacillus apinorum and Klebsiella pneumoniae for non-PDO honeys. In addition, potential microbial biomarkers of some honey varieties including L. kunkeei for Albaida and Retama del Teide varieties, and P. apium for Tajinaste variety, were identified. ML models were validated on an independent set of samples leading to high accuracy rates (above 90 %). This work demonstrates the potential of ML to differentiate different types of honey using metagenome-based methods, leading to high performance metrics. In addition, ML models discriminate both the geographical origin and variety of products corresponding to different PDOs and non-PDO products. Results here presented may contribute to develop enhanced traceability and authenticity methods that could be applied to a wide range of foods.

Using pollen DNA metabarcoding to trace the geographical and botanical origin of honey from Karangasem, Indonesia

The unique floral fingerprint embedded within honey holds valuable clues to its geographical and botanical origin, playing a crucial role in ensuring authenticity and detecting adulteration. Honey from native Apis cerana and Heterotrigona itama bees in Karangasem, Indonesia, was examined utilizing pollen DNA metabarcoding for honey source identification. In this study, we used ITS2 amplicon sequencing to identify floral DNA in honey samples. The finding reveals distinct pollen signatures for each bee species. Results analysis showed A. cerana honey generated 179,267 sequence reads, assembled into Amplicon Sequence Variants (ASVs) with a total size of 485,932 bp and an average GC content of 59 %. H. itama honey generated 177,864 sequence reads, assembled into ASVs with a total size of 350,604 bp and an average GC content of 57 %. A. cerana honey exhibited a rich tapestry of pollen from eleven diverse genera, with Schleichera genus dominating at an impressive relative read abundance of 72.8 %. In contrast, H. itama honey displayed a remarkable mono-dominance of the Syzygium genus, accounting for a staggering 99.95 % of its pollen composition or relative read abundance, highlighting their distinct foraging preferences and floral resource utilization. Notably, all identified pollen taxa were indigenous to Karangasem, solidifying the geographical link between honey and its origin. This study demonstrates pollen DNA metabarcoding may identify honey floral sources. By using pollen profiles from different bee species and their foraging patterns, we may protect consumers against honey adulteration and promote sustainable beekeeping in Karangasem district. Future research could explore expanding the database of reference pollen sequences and investigating the influence of environmental factors on pollen composition in honey. Investigating this technology's economic and social effects on beekeepers and consumers may help promote fair trade and sustainable beekeeping worldwide.

An updated review of functional ingredients of Manuka honey and their value-added innovations

Manuka honey (MH) is a highly prized natural product from the nectar of Leptospermum scoparium flowers. Increased competition on the global market drives MH product innovations. This review updates comparative and non-comparative studies to highlight nutritional, therapeutic, bioengineering, and cosmetic values of MH. MH is a good source of phenolics and unique chemical compounds, such as methylglyoxal, dihydroxyacetone, leptosperin glyoxal, methylsyringate and leptosin. Based on the evidence from in vitro, in vivo and clinical studies, multifunctional bioactive compounds of MH have exhibited anti-oxidative, anti-inflammatory, immunomodulatory, anti-microbial, and anti-cancer activities. There are controversial topics related to MH, such as MH grading, safety/efficacy, implied benefits, and maximum levels of contaminants concerned. Artificial intelligence can optimize MH studies related to chemical analysis, toxicity prediction, multi-functional mechanism exploration and product innovation.

An innovative molecular approach towards the cost-effective entomological authentication of honey

Honey authentication and traceability are crucial not only for economic purposes but also for ensuring safety. However, the widespread adoption of cutting-edge technologies in practical applications has been hampered by complex, time-consuming sample pre-treatment processes, the need for skilled personnel, and substantial associated expenses. This study aimed to develop a simple and cost-effective molecular technique to verify the entomological source of honey. By utilizing newly designed primers, we successfully amplified the mitochondrial 16S ribosomal RNA gene of honey bees from honey, confirming the high quality of the extracted DNA. Employing RFLP analysis with AseI endonuclease, species-specific restriction patterns were generated for honey derived from six closely related honey bees of the Apis genus. Remarkably, this method was proven equally effective in identifying heat-treated and aged honey by presenting the same RFLP profiles as raw honey. As far as we know, this is the initial research of the simultaneous differentiation of honey from closely related honey bee species using the restriction endonuclease AseI and mitochondrial 16S rRNA gene fragments. As a result, it holds tremendous potential as a standardized guideline for regulatory agencies to ascertain the insect origins of honey and achieve comprehensive traceability.

The buzz about honey-based biosurveys

Approximately 1.8 million metric tonnes of honey are produced globally every year. The key source behind this output, the honey bee (Apis mellifera), works tirelessly to create the delicious condiment that is consumed worldwide. The honey that finds its way into jars on store shelves contains a myriad of information about its biogeographical origins, such as the bees that produced it, the botanical constituents, and traces of other organisms or pathogens that have come in contact with the product or its producer. With the ongoing threat of honey bee decline and overall global biodiversity loss, access to ecological information has become an key factor in preventing the loss of species. This review delves into the various molecular techniques developed to characterize the collective DNA harnessed within honey samples, and how it can be used to elucidate the ecological interactions between honey bees and the environment. We also explore how these DNA-based methods can be used for large-scale biogeographical studies through the environmental DNA collected by foraging honey bees. Further development of these techniques can assist in the conservation of biodiversity by detecting ecosystem perturbations, with the potential to be expanded towards other critical flying pollinators.

Combined quantitative lipidomics and back-propagation neural network approach to discriminate the breed and part source of lamb

The study successfully utilized an analytical approach that combined quantitative lipidomics with back-propagation neural networks to identify breed and part source of lamb using small-scale samples. 1230 molecules across 29 lipid classes were identified in longissimus dorsi and knuckle meat of both Tan sheep and Bahan crossbreed sheep. Applying multivariate statistical methods, 12 and 7 lipid molecules were identified as potential markers for breed and part identification, respectively. Stepwise linear discriminant analysis was applied to select 3 and 4 lipid molecules, respectively, for discriminating lamb breed and part sources, achieving correct rates of discrimination of 100 % and 95 %. Additionally, back-propagation neural network proved to be a superior method for identifying sources of lamb meat compared to other machine learning approaches. These findings indicate that integrating lipidomics with back-propagation neural network approach can provide an effective strategy to trace and certify lamb products, ensuring their quality and protecting consumer rights.

A preliminary metabarcoding analysis of Portuguese raw honeys

The microbial diversity in Portuguese raw honeys remains largely uncharacterized, constituting a serious knowledge gap in one of the country's most important resources. This work provides an initial investigation with amplicon metabarcoding analysis of two Lavandula spp. from different geographical regions of Portugal and one Eucalyptus spp. honey. The results obtained allowed to identify that each honey harbors diverse microbiomes with taxa that can potentially affect bee and human health, cause spoilage, and highlight bad bee-hive management practices. We verified that prokaryotes had a tendency towards a more marked core bacterial and a relative homogenous taxa distribution, and that the botanical origin of honey is likely to have a stronger impact on the fungal community. Thus, the results obtained in this work provide important information that can be helpful to improve this critical Portuguese product and industry.

A comprehensive review of the current trends and recent advancements on the authenticity of honey

The authenticity of honey currently poses challenges to food quality control, thus requiring continuous modernization and improvement of related analytical methodologies. This review provides a comprehensively overview of honey authenticity challenges and related analytical methods. Firstly, direct and indirect methods of honey adulteration were described in detail, commenting the existing challenges in current detection methods and market supervision approaches. As an important part, the integrated metabolomic workflow involving sample processing procedures, instrumental analysis techniques, and chemometric tools in honey authenticity studies were discussed, with a focus on their advantages, disadvantages, and scopes. Among them, various improved microscale extraction methods, combined with hyphenated instrumental analysis techniques and chemometric data processing tools, have broad application potential in honey authenticity research. The future of honey authenticity determination will involve the use of simplified and portable methods, which will enable on-site rapid detection and transfer detection technologies from the laboratory to the industry.

A primer on pollen assignment by nanopore-based DNA sequencing

The possibility to identify plants based on the taxonomic information coming from their pollen grains offers many applications within various biological disciplines. In the past and depending on the application or research in question, pollen origin was analyzed by microscopy, usually preceded by chemical treatment methods. This procedure for identification of pollen grains is both time-consuming and requires expert knowledge of morphological features. Additionally, these microscopically recognizable features usually have a low resolution at species-level. Since a few decades, DNA has been used for the identification of pollen taxa, as sequencing technologies evolved both in their handling and affordability. We discuss advantages and challenges of pollen DNA analyses compared to traditional methods. With readers with little experience in this field in mind, we present a hands-on primer for genetic pollen analysis by nanopore sequencing. As our lab mainly works with pollen collected within agroecological research projects, we focus on pollen collected by pollinating insects. We briefly consider sample collection, storage and processing in the laboratory as well as bioinformatic aspects. Currently, pollen metabarcoding is mostly conducted with next-generation sequencing methods that generate short sequence reads (<1 kb). Increasingly, however, pollen DNA analysis is carried out using the long-read generating (several kb), low-budget and mobile MinION nanopore sequencing platform by Oxford Nanopore Technologies. Therefore, we are focusing on aspects for palynology with the MinION DNA sequencing device.

Towards DNA-Based Methods Analysis for Honey: An Update

Honey is a natural product widely consumed all over the world due to its relationship with healthy benefits. Additionally, environmental and ethical issues have a higher weight in the consumer's choice to buy honey as a natural product. Following the high demand of this product, several approaches have been suggested and developed aiming at the assessment of honey's quality and authenticity. Target approaches, such as pollen analysis, phenolic compounds, sugars, volatile compounds, organic acids, proteins, amino acids, minerals, and trace elements, showed an efficacy, particularly concerning the honey origin. However, a special highlight is given to DNA markers, due to their useful applicability in environmental and biodiversity studies, besides the geographical, botanical, and entomological origins. Different DNA target genes were already explored for addressing diverse sources of honey DNA, with DNA metabarcoding attaining a relevant importance. This review aims to describe the latest advances on DNA-based methods applied in honey related studies, identifying the research needs for the development of new and additional required methodologies, and to select the most adequate tools for future research projects.

An Accurate and Rapid Way for Identifying Food Geographical Origin and Authenticity: Editable DNA-Traceable Barcode

DNA offers significant advantages in information density, durability, and replication efficiency compared with information labeling solutions using electronic, magnetic, or optical devices. Synthetic DNA containing specific information via gene editing techniques is a promising identifying approach. We developed a new traceability approach to convert traditional digitized information into DNA sequence information. We used encapsulation to make it stable for storage and to enable reading and detection by DNA sequencing and PCR-capillary electrophoresis (PCR-CE). The synthesized fragment consisted of a short fragment of the mitochondrial cytochrome oxidase subunit I (COI) gene from the Holothuria fuscogilva (ID: LC593268.1), inserted geographical origin information (18 bp), and authenticity information from Citrus sinensis (20 bp). The obtained DNA-traceable barcodes were cloned into vector PMD19-T. Sanger sequencing of the DNA-traceable barcode vector was 100% accurate and provided a complete readout of the traceability information. Using selected recognition primers CAI-B, DNA-traceable barcodes were identified rapidly by PCR amplification. We encapsulated the DNA-traceable barcodes into amorphous silica spheres and improved the encapsulation procedure to ensure the durability of the DNA-traceable barcodes. To demonstrate the applicability of DNA-traceable barcodes as product labels, we selected Citrus sinensis as an example. We found that the recovered and purified DNA-traceable barcode can be analyzed by standard techniques (PCR-CE for DNA-traceable barcode identification and DNA sequencing for readout). This study provides an accurate and rapid approach to identifying and certifying products' authenticity and traceability.

Micro"bee"ota: Honey Bee Normal Microbiota as a Part of Superorganism

Honey bees are model organisms for microbiota research. Gut microbiomes are very interesting for surveys due to their simple structure and relationship with hive production. Long-term studies reveal the gut microbiota patterns of various hive members, as well as the functions, sources, and interactions of the majority of its bacteria. But the fungal non-pathogenic part of gut microbiota is almost unexplored, likewise some other related microbiota. Honey bees, as superorganisms, interact with their own microorganisms, the microbial communities of food stores, hive surfaces, and other environments. Understanding microbiota diversity, its transition ways, and hive niche colonization control are necessary for understanding any separate microbiota niche because of their interplay. The long coevolution of bees with the microorganisms populating these niches makes these systems co-dependent, integrated, and stable. Interaction with the environment, hive, and other bees determines caste lifestyle as well as individual microbiota. In this article, we bring together studies on the microbiota of the western honey bee. We show a possible relationship between caste determination and microbiota composition. And what is primary: caste differentiation or microbiota composition?

Significant compositional and functional variation reveals the patterns of gut microbiota evolution among the widespread Asian honeybee populations

The gut microbiome is a crucial element that facilitates a host’s adaptation to a changing environment. Compared to the western honeybee Apis mellifera, the Asian honeybee, Apis cerana populations across its natural range remain mostly semi-feral and are less affected by bee management, which provides a good system to investigate how gut microbiota evolve under environmental heterogeneity on large geographic scales. We compared and analyzed the gut microbiomes of 99 Asian honeybees, from genetically diverged populations covering 13 provinces across China. Bacterial composition varied significantly across populations at phylotype, sequence-discrete population (SDP), and strain levels, but with extensive overlaps, indicating that the diversity of microbial community among A. cerana populations is driven by nestedness. Pollen diets were significantly correlated with both the composition and function of the gut microbiome. Core bacteria, Gilliamella and Lactobacillus Firm-5, showed antagonistic turnovers and contributed to the enrichment in carbohydrate transport and metabolism. By feeding and inoculation bioassays, we confirmed that the variations in pollen polysaccharide composition contributed to the trade-off of these core bacteria. Progressive change, i.e., nestedness, is the foundation of gut microbiome evolution among the Asian honeybee. Such a transition during the co-diversification of gut microbiomes is affected by environmental factors, diets in general, and pollen polysaccharides in particular.

Physicochemical Properties of a New Green Honey from Banggi Island, Sabah

Green honey is exclusively available on the island of Banggi in Sabah, and its uniqueness sees the commodity being sold at a high market price. Therefore, green honey is prone to adulteration by unscrupulous individuals, possibly compromising the health of those consuming this food commodity for its curative properties. Moreover, an established standard for reducing sugar in green honey is unavailable. Ipso facto, the study aimed to profile green honey’s physical and chemical properties, such as its pH, moisture content, free acidity, ash content, electroconductivity, hydroxymethylfurfural (HMF), total phenolic content, total flavonoid content, DPPH, colour, total sugar content, total protein content, and heavy metals as well as volatile organic compounds, the data of which are profoundly valuable in safeguarding consumers’ safety while providing information for its quality certification for local consumption and export. The results revealed that the honey’s physicochemical profile is comparable to other reported kinds of honey. The honey’s naturally green colour is because of the chlorophyll from the nectar from various flowers on the island. The raw honey showed free acidity between 28 and 33 Meq/100 g, lower than the standard’s 50 Meq/100 g. The hydroxymethylfurfural content is the lowest compared to other reported honey samples, with the total phenolic content between 16 and 19 mg GAE/100 g. The honey’s reducing sugar content is lower (~37.9%) than processed ones (56.3%) because of water removal. The protein content ranged from 1 to 2 gm/kg, 4- to 6-fold and 2-fold higher than local and manuka honey, respectively. The exceptionally high content of trans-4-hydroxyproline in raw honey is its source of collagen and other healing agents. Interestingly, low levels of arsenic, lead, nickel, cadmium, copper, and cobalt were detected in the honey samples, presumably due to their subterranean hives. Nevertheless, the honey is fit for general consumption as the concentrations were below the maxima in the Codex Alimentarius Commission of 2001.