Genetic profile-based targeted crop protection product application

Abstract

Systems, methods, and computer programs disclosed herein relate to determining the efficacy of one or more crop protection products against pests based on genetic profiles.

Description

Technical Field

[0001] The systems, methods, and computer programs disclosed herein relate to determining the efficacy of one or more crop protection products against pests based on genetic profiles. Background Art

[0002] Ensuring global food security requires investing in sustainable agricultural production of crops. Harmful organisms (such as plant pathogenic fungi, insects, and bacteria) are a significant risk in every vegetation phase of food production. Strategies for preventing and controlling these harmful organisms include the use of crop protection products, but harmful organisms continue to evolve, becoming more aggressive and more adaptable to their corresponding hosts. In particular, plant pathogenic fungi and insects can evade the effects of crop protection products through genetic mutations, which occur not only in the targeted genes targeted by the crop protection products, but also in other genes involved in pesticide metabolism or detoxification. This mechanism is known as pesticide resistance. Different factors play a role in the formation of this resistance, such as agronomic methods, the harmful organisms involved, the rate and timing of application of crop protection products, and the affected crops. At the same time, efforts are being made to reduce the amount of chemical crop protection products used in agriculture to minimize the impact on the environment. Therefore, it is necessary to develop methods that allow for highly customized, automated, and precise application of crop protection products, particularly in a site-specific manner, to prevent and control harmful organisms and minimize the formation of pesticide resistance.

[0003] WO2021 / 228578A1 describes a method for identifying variants in a site-specific manner in the field, determining their sequence information, and comparing this information with known resistance markers or predicted structures (such as the α-sheet) to estimate the effect of the variant on inhibitor binding. Using DNA or RNA analysis to identify resistance allows for very early detection of resistant pests in the field.

[0004] In WO 2022 / 099169 A1 a method is described which uses pesticide resistance factors and geographic location to correlate the frequency of certain genotypes with resistance to generate a map.

[0005] However, information on the presence of resistant pests in a field does not include any information on how resistance affects the efficacy of crop protection products, nor does it include any information on selecting alternative crop protection products that may provide better efficacy.

[0006] Furthermore, information about the presence of resistant pests in a farmer's fields does not provide the farmer with information about how effectively he or she can control the pest despite the presence of resistance. Summary of the Invention

[0007] These and additional problems are addressed by the present disclosure.

[0008] Still further, the present disclosure does not merely analyze sequence information from one specific variant, but instead uses a model in which the genetic profile of one or more variants is correlated with efficacy data for one or more crop protection products and optionally other agricultural data.

[0009] Thus, in a first aspect, the present disclosure relates to a computer-implemented method comprising:

[0010] - receiving a genetic profile of one or more pests from an agricultural or horticultural area, the genetic profile comprising one or more variants;

[0011] - identifying one or more biological targets based on the genetic profile;

[0012] - Identify one or more active ingredients that act on one or more biological targets;

[0013] - determining the expected efficacy of a crop protection product comprising one or more active ingredients using a model in which, for a large number of crop protection products, their efficacy is correlated with one or more genetic profiles of one or more pests;

[0014] -Deliver the desired efficacy of crop protection products.

[0015] In another aspect, the present disclosure provides a computer system comprising:

[0016] processor; and

[0017] A memory storing an application program, wherein the application program is configured to perform an operation when executed by the processor, the operation comprising:

[0018] - receiving a genetic profile of one or more pests from an agricultural or horticultural area, said genetic profile comprising one or more variants;

[0019] - identifying one or more biological targets based on the genetic profile;

[0020] - Identify one or more active ingredients that act on one or more biological targets;

[0021] - determining the expected efficacy of a crop protection product comprising one or more active ingredients using a model in which, for a large number of crop protection products, their efficacy is correlated with one or more genetic profiles of one or more pests;

[0022] -Deliver the desired efficacy of crop protection products.

[0023] In another aspect, the present disclosure provides a non-transitory computer-readable storage medium having software instructions stored thereon that, when executed by a processor of a computer system, cause the computer system to perform the following steps:

[0024] - receiving a genetic profile of one or more pests from an agricultural or horticultural area, said genetic profile comprising one or more variants;

[0025] - identifying one or more biological targets based on the genetic profile;

[0026] - Identify one or more active ingredients that act on one or more biological targets;

[0027] - determining the expected efficacy of a crop protection product comprising one or more active ingredients using a model in which, for a large number of crop protection products, their efficacy is correlated with one or more genetic profiles of one or more pests;

[0028] -Deliver the desired efficacy of crop protection products.

[0029] Further aspects and preferred embodiments are defined in the dependent claims, the description and the drawings. DETAILED DESCRIPTION

[0030] The present invention will be explained in more detail below without distinguishing between the various aspects of the disclosure (method, computer system, computer-readable storage medium). On the contrary, the following explanation is intended to apply in an analogous manner to all aspects of the disclosure, regardless of the context in which they appear (method, computer system, computer-readable storage medium).

[0031] If steps are stated in a certain order in this description or in the claims, this does not necessarily mean that the invention is limited to the stated order. On the contrary, it is conceivable that these steps can also be performed in a different order or in parallel with each other, unless one step builds on another step, which necessarily requires the building step to be performed later (however, this is clear in individual cases). The stated order is therefore a preferred embodiment of the present disclosure.

[0032] As used herein, the articles "a" and "an" are intended to include one or more items and are used interchangeably with "one or more" and "at least one." As used in the specification and claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. When only one item is intended, the term "one" or similar terms is used. Furthermore, as used herein, the terms "has," "have," "having," and the like are intended to be open-ended terms.

[0033] Further, the phrase "based on" means "based, at least in part, on" unless explicitly stated otherwise.

[0034] Some embodiments of the present disclosure will be described more fully below with reference to the accompanying drawings, which illustrate some, but not all, embodiments of the present disclosure. Indeed, the various embodiments of the present disclosure may be embodied in many different forms and should not be construed as limited to the embodiments described herein; rather, these example embodiments are provided so that this disclosure will be thorough and complete and will fully convey the scope of the disclosure to those skilled in the art.

[0035] The present disclosure provides means for determining the efficacy of one or more crop protection products against one or more pests.

[0036] The term "crop protection product" refers to a composition which is used to protect plants or plant products from pests or to prevent such exposure, to destroy unwanted plants or plant parts and / or to inhibit the growth of unwanted plants or to prevent such growth. Examples of crop protection products are herbicides, insecticides, nematicides, acaricides or fungicides. Crop protection products typically contain one or more active ingredients.

[0037] An "active ingredient" is a chemical or biological substance that has a specific effect or causes a specific response in an organism.

[0038] "Harmful organisms" refer to organisms that are pathogenic organisms or disease transmitters in humans or animals, or that can damage crop plants, adversely affect the harvest of crop plants, or compete with crop plants for natural resources. Examples of harmful organisms are disease vectors, broadleaf weeds, grass weeds, animal pests (such as beetles, mites, spiders, caterpillars, nematodes, arachnids, snails, slugs and worms) and pathogenic microorganisms (such as fungi, bacteria, oomycetes and viruses). Although biologically speaking, viruses are not counted as organisms, it is currently intended to include them in the entry for harmful organisms. Harmful organisms may also be weeds. In one embodiment, a species of harmful organisms is provided or received. In another embodiment, a strain, cultivar or subtype of a species is provided or received. The strain, cultivar or subtype represents a genetic variant within the species.

[0039] The term "weed" (plural: weeds) refers to plants of the spontaneous vegetation (grain flora) in plots of crop plants, on meadows or in gardens, which are not intentionally planted in these environments by humans, but instead grow, for example, from latent or airborne seeds in the soil. Strictly speaking, this term is not restricted to broadleaf plants, but also includes grasses, ferns, mosses or woody plants. In the field of crop protection, another frequently used term is "gramineous weed" (plural: gramineous weeds) to clarify the distinction with herbaceous plants. In this document, the term "weed" is used as a general term, which is intended to also include the term grass weeds, unless specific broadleaf weeds or grass weeds are mentioned.

[0040] The term "efficacy" refers to the ability to produce a desired or expected result. For the purposes of this disclosure, "efficacy" preferably refers to the ability of an active ingredient or crop protection product to exert an effect on a pest (i.e., weaken, kill, or render the pest harmless). Efficacy can be measured by any method suitable for determining the effect of an active ingredient and / or crop protection product, including identifying an appropriate dosage. Efficacy can be measured as an EC50 value, which is the concentration of the active ingredient or crop protection product at which 50% of the activity of a biological target in a pest is inhibited in a corresponding assay or test.

[0041] "Biological target" means an enzyme or protein in a pest that a) has its function altered as a result of interaction with a crop protection product or one or more active ingredients, thereby negatively affecting the organism, or b) is capable of altering the crop protection product or one or more active ingredients so that they are no longer active, or is capable of removing the crop protection product or one or more active ingredients from the pest.

[0042] In a first step of a method according to the present disclosure, a genetic profile of one or more pests from an agricultural or horticultural area is received.

[0043] An "agricultural or horticultural area" is a spatially definable area on the Earth's surface where crops are grown. Preferably, this area is at least partially used for agricultural purposes, in that the crop plants are grown in one or more fields and / or greenhouses, fed, and harvested. This area may also be or include areas of the Earth's surface used for forestry purposes (e.g., forests). Gardens, parks, and other areas covered with vegetation used solely for human enjoyment also belong to "agricultural or horticultural areas."

[0044] A "genetic profile" is the result of a quantitative or qualitative analysis of variants in samples taken from an agricultural or horticultural area. A genetic profile provides a list of variants present in those samples, and optionally, sequence information for the corresponding variants and the percentage of the variant in the sample.

[0045] " Variant " refers to a specific isomer of a haplotype or polymorphism in a population of a pest, and the difference between this specific form and other forms of the same haplotype is that at least one (usually more than one) variant site or sequence of nucleotides within the gene sequence. The different sequences at these variant sites between different alleles of a gene are referred to as "gene sequence variants", "alleles", "differences" or "variants". An allele can be distinguished from other alleles by having at least one (usually more than one) variant site within the gene sequence. Other terms equivalent to "variant" known in the art include mutation and single nucleotide polymorphism (SNPs). The presence of one or more variants mentioned refers to specific variants, i.e., specific nucleotides at specific polymorphic sites, and not just any variants present in the gene. Variant is an embodiment of genetic information. In one embodiment, variant relates to genetic information describing the resistance of an organism to a crop protection product.

[0046] Genetic information refers to any information about the genetic characteristics of an organism, including but not limited to DNA, sequences, RNA sequences, partial DNA and / or RNA sequences, molecular structures of DNA or RNA, epigenetic information (e.g., methylation of a portion of DNA), gene mutation information, gene copy number variation information, gene overexpression information, gene expression level information, gene transfer information, information about the ratio between wild type and mutants, information about the ratio between different mutants, information about the ratio between mutants and other variants (e.g., epigenetic variants), information about the ratio of different variants (e.g., epigenetic variants), information about the types of plant diseases (e.g., Septoria, yellow rust, Asian soybean rust) or other diseases. In the context of the present invention, the term "genetic information" also includes information about the lack of certain wild types, mutants or variants (e.g., epigenetic variants) or DNA / RNA sequences, or partial DNA / RNA sequences, or specific epigenetic information. In the context of the present invention, the term "genetic information" also includes information about the lack of specific genetic information (e.g., information about the lack of a specific type of Septoria is also genetic information). In a preferred embodiment of the present invention, "genetic information" is at least one of the following information: DNA sequence, RNA sequence, partial DNA and / or RNA sequence, molecular structure of DNA and / or RNA, epigenetic information (such as methylation of partial DNA), gene mutation information, gene copy number variation information, gene overexpression information, gene expression level information, gene transfer information, information about the ratio between wild type and mutant type, information about the ratio between different mutants, information about the ratio between mutants and other variants (e.g., epigenetic variants), information about the ratio between different variants (e.g., epigenetic variants), information about the type of plant disease (e.g., Septoria, yellow rust, Asian soybean rust) or other diseases. In another preferred embodiment of the present invention, "genetic information" is at least one of the following information: DNA sequence, RNA sequence, molecular structure of DNA and / or RNA, partial DNA and / or RNA sequence, epigenetic information (e.g., methylation of partial DNA). In another preferred embodiment of the present invention, "genetic information" is at least one of the following information: DNA sequence, RNA sequence.In another preferred embodiment of the present invention, "genetic information" is at least one of the following information: gene mutation information, gene copy number variation information, gene overexpression information, gene expression level information, gene transfer information, information on the ratio between wild type and mutant, information on the ratio between different mutants, information on the ratio between mutants and other variants (e.g., epigenetic variants), information on the ratio between different mutants (e.g., epigenetic variants), information on the type of plant disease (e.g., Septoria, yellow rust, Asian soybean rust) or other diseases. In another preferred embodiment of the present invention, "genetic information" is at least one of the following information: gene mutation information, gene copy number variation information, gene overexpression information, gene expression level information, gene transfer information. In another preferred embodiment of the present invention, "genetic information" is at least one of the following information: information on the ratio between wild type and mutant, information on the ratio between different mutants, information on the ratio between mutants and other variants (e.g., epigenetic variants), information on the ratio between different variants (e.g., epigenetic variants).

[0047] As used herein, "isoform" refers to a specific form of a gene, mRNA, cDNA, or protein encoded thereby, which is distinguished from other forms by a specific sequence and / or structure.

[0048] "Haplotype" refers to a genetic variant or variant combination carried on at least one chromosome of a strain of a pest. A haplotype typically includes a plurality of adjacent polymorphic loci. All parts of a haplotype as used herein appear on the identical replica of a chromosome or haploid DNA molecule. In the absence of contrary evidence, a haplotype is considered to be a combination of a plurality of loci that may be transmitted together during meiosis. Each pest carries at least one haplotype for any given genetic locus, which includes sequences on the homologous chromosomes inherited from parents. For a given locus, these haplotypes can be identical or can represent one or more different genetic variants. Haplotype analysis is the process of determining one or more haplotypes in an individual. Haplotype analysis can include the use of family pedigrees, molecular techniques, and / or statistical inferences.

[0049] "Polymorphism" refers to the presence of two or more different nucleotide sequences on a specific locus of genomic DNA. Polymorphism can be used as a genetic marker or as a variant. Polymorphisms include nucleotide substitutions, insertions, deletions, and microsatellites, and may (but not necessarily) produce detectable differences in gene expression or protein function. A polymorphic site refers to a nucleotide position within the locus of at least one individual in a population at a nucleotide sequence different from a reference sequence. As used herein, "deletion / insertion polymorphism" or "DIP" is the insertion of one or more nucleotides in one version of a sequence relative to another version. If it is known which alleles represent minor alleles, the term "deletion" is used when the minor allele is a deleted nucleotide, and the term "insertion" is used when the minor allele is an added nucleotide. When there are multiple forms or lengths and the minor allele is not obvious, the term "deletion / insertion polymorphism" is also used. For example, for poly-T polymorphisms described herein, polymorphisms of multiple lengths can be observed.

[0050] As used herein, "genetic locus" or "locus" refers to a position on a chromosome or DNA molecule that generally corresponds to a gene or a physical or phenotypic characteristic, or to a specific nucleotide or nucleotide segment. Loci is the plural form of locus.

[0051] The genetic profile is preferably in the form of digital data."Digital" means data that can be processed by a machine (eg, a computer system).

[0052] The term "receiving data" refers to accepting, retrieving, and / or obtaining data from one or more sources. A genetic profile can be entered into a computer system of the present disclosure by a user thereof (e.g., a farmer), and / or can be read from one or more data storage devices, and / or can be transmitted from one or more separate computer systems and / or devices (e.g., a device used for sequencing genes in a sample).

[0053] In order to generate a genetic profile in a first step, samples have to be collected from an agricultural or horticultural area.

[0054] The sample may include soil, water, air, plants, seedlings or plant parts (such as leaves, stems, flowers, roots, seeds), or weeds, or samples infected or at risk of infection by harmful organisms (such as pests, fungi, bacteria, oomycetes or viruses).

[0055] The term "collect" should not be construed as limiting in any way. An example of a synonym is the term "sampling".

[0056] The sampling may be automated. "Automated" means that the sampling is performed by a machine or machines without human input. Similarly, the steps of determining geographic coordinates, processing samples, sequencing, analyzing DNA and / or RNA sequences, and / or entering information about resistance into a resistance map may also be automated.

[0057] Sampling can be performed using a portable device carried by a person, or using a vehicle that moves or is moved within and / or over a field of crop plants. For example, it is conceivable to use (preferably unmanned) land machines and / or (preferably unmanned) aircraft (e.g., drones) and / or robots. Alternatively, sampling can be performed by one or more devices that are configured in a fixed manner at a certain location. Furthermore, it is conceivable that a user carries a mobile device, performs the sampling themselves, and provides the sample to the device.

[0058] The nature of the sample depends on the pest and its purpose is to check for individuals that are developing or have developed resistance to one or more crop protection products. The sample includes one or more pests, or parts of one or more pests.

[0059] Sampling can be done with the aid of image recognition technology. It is conceivable that, for example, a camera generates digital images of plants in a field and transmits these images to an image analysis unit. The image analysis unit is configured to identify features in the image that indicate the presence of pests, such as broadleaf / grass weeds and / or crops infested by one or more pests. The images can be analyzed using, for example, pattern recognition technology or other self-learning systems (e.g., artificial neural networks). It is conceivable that the image analysis unit is configured to identify one or more defined grass / broadleaf weed species and / or symptoms caused by one or more pests. It is also conceivable that the image analysis unit is configured to identify that the plants in the image are not crop plants being grown (and therefore may be plants that compete with crop plants for resources and / or affect the quality of the harvest). Methods and systems for identifying broadleaf / grass weeds and / or crops infested by one or more pests are described in the prior art (see, for example, WO 2017 / 194398 A1, WO 2017 / 194399 A1). However, it is also conceivable to take samples from a plant that one might consider to be a broadleaf / grass weed because, for example, it grows where no crop plants are planted / sown, or it has reached another stage of development compared to the crop plants planted / sown in the field.

[0060] If the harmful organism is a fungus, virus, oomycete, or bacterium, the sample is preferably taken from the infecting organism (the fungus, virus, or bacterium is located in / on the infecting organism). The infecting organism can be identified, for example, using image recognition technology; for example, Peat GmbH commercially offers a software application ("app") for identifying plant diseases based on image recognition technology (https: / / plantix.net / de).

[0061] If the pest is an animal pest, such as an insect (at various stages, from egg, larva, caterpillar, pseudocaterpillar to adult), a slug or snail, a worm (nematode), or an arachnid, a trap can be used to capture the animal pest and then provide the pest (or its parts) for analysis. There are various options for capturing animal pests, such as rubberized plates, pan traps (e.g., pans filled with water and a surfactant), etc. The trap can be baited to attract the animal pest.

[0062] In one embodiment, the samples are taken from air, water and / or soil samples where the pests are located. It is also conceivable that animal excreta are sampled.

[0063] In one embodiment, the reason for sampling is that the pest has been observed, for example by inspection or automatically by image recognition technology.

[0064] In one embodiment, sampling is performed due to a suspected incidence of a pest. For example, it is conceivable that, using a predictive model, a risk of infestation by a pest has been determined to be above a defined threshold. It is conceivable that an infestation has been observed near the location where the sample was collected.

[0065] In one embodiment, the reason for sampling is the suspicion of the presence or impending development of resistance. It is conceivable, for example, that when using a crop protection product for combating pests, it is observed that the crop protection product does not develop the desired effect.

[0066] In one embodiment, one or more samples may be obtained at a single location. In one embodiment, one or more samples may be obtained at more than one location.

[0067] For those locations where one or more samples have been taken, the geographical coordinates associated with that location are identified. This is important so that information about resistance can be associated with the respective location and entered into the resistance map.

[0068] Geographic coordinates are typically determined using a positioning system. One known positioning system is a satellite navigation system, such as NAVSTAR GPS, GLONASS, Galileo, or BeiDou. Given that the abbreviation GPS (Global Positioning System) is well established in everyday language as a generic name for all satellite navigation systems, the term GPS will be used hereinafter as a general term for all positioning systems.

[0069] In one embodiment, the sampling device has a GPS sensor. The mobile device can be moved, for example, through a field of crop plants, and one or more samples can be obtained at a plurality of locations. The GPS sensor can be used to confirm the geographical coordinates of the locations where the samples have been obtained, and this information can be stored in the device's data memory and / or transmitted to an external computer system via a (mobile) network. It is also conceivable that the mobile device uses the GPS sensor to travel to one or more locations defined in advance. For example, a prediction model may have identified one or more locations with an increased risk of (resistant) pest incidence. These locations can be found by the mobile device, thereby obtaining one or more samples at these locations. When the samples have been obtained and analyzed, the results of the analysis can be stored in the device's data memory together with the geographical coordinates and / or transmitted to an external computer system via a (mobile) network.

[0070] In one embodiment, the mobile device can be moved by the user. At the location where the user obtains the sample, the geographic coordinates of the sampling location are confirmed / recorded. It is conceivable that the user can be guided to a predefined location via the mobile device with the help of a GPS sensor. In that case, the user can be assisted by augmented reality technology. Using these technologies, the real world is displayed on a screen, and this display is optically extended by additional computer-generated information fragments. It is conceivable, for example, to display the real world on the screen of a computer system (e.g., a smartphone) in so-called real-time mode, and to enhance this display with virtual objects representing the traps that have been set and the traps to be found. It is also conceivable to use a head-up display or a head-mounted display (e.g., video glasses (Eye Tap)). It is also conceivable that the user is guided to a pre-set trap for pests, and the guidance is performed using a GPS sensor and optional augmented reality technology.

[0071] If the device for sampling is a fixed unit, it can also be equipped with a GPS sensor so that the device location can be recorded / confirmed. However, preferably, the device location has been determined or confirmed when the device is installed. For example, it is possible to install the device in a location and then use a (separate) GPS sensor to determine the geographical coordinates of the location where the device is installed. It is also conceivable to determine the geographical coordinates of the location where the device is to be installed and then install the device at the determined location accordingly. The situation of the device can be marked / has been marked, for example, in a database and / or on a digital map. It is possible that the device has a clear identifier (for example, an ID number). When this device transmits information about the sample to an external computer system, it authenticates itself by, for example, a clear identifier. The database can include a record of the location of a specific device with a specific identifier that has been installed. Therefore, by querying the database using a clear identifier, the location of the device can also be confirmed.

[0072] The advantage of using global satellite navigation systems to determine location is high accuracy. An alternative method (better but less accurate) uses radio standards for location determination. Positioning can be performed using mobile phones (cell phones). In mobile radio, the simplest type of location determination is based on the fact that the cell in which the mobile phone is located is known. Because, for example, a switched-on mobile phone communicates with a base station, the mobile phone's position can be assigned to at least one mobile cell (cell ID). Using GSM (Global System for Mobile Communications), the location of a transmitting unit can be determined to an accuracy of several hundred meters. In cities, the location can be determined to an accuracy of 100 to 500 meters; in rural areas, the radius increases to 10 kilometers or more. Accuracy can be further improved if the information provided by the cell ID is combined with the TA parameter (TA: Timing Advance). The larger this value, the farther the transmitting unit is from the base station. Using the EOTD method (EOTD: Enhanced Observed Time Difference), a transmitting unit can be located with much higher accuracy. In this method, the difference in the transmission time of the signal between the transmitting unit and several receiving units is determined.

[0073] In one embodiment, the transmission of information and location determination is performed via the Sigfox network. Sigfox is a low-power wide area network (LPWAN) designed for small data packets and very energy-efficient operation. Sigfox base stations are able to communicate over long distances without being affected by interference. The range of a single base station (which can manage up to 1 million transmission units) is 3 km to 5 km in densely populated centers and 30 km to 70 km in rural areas. With Sigfox, data packets are received from all base stations within the transmission area. In this way, the location of the transmission unit can be determined.

[0074] The geographical coordinates are preferably confirmed with an accuracy of at least 100 meters. In one embodiment, the geographical coordinates are preferably confirmed with an accuracy of at least 1, 2, 5, 10, 20, 25, 50 or 75 meters.

[0075] After having obtained the sample, the sample is processed. This can be done by isolating DNA from a crude or purified sample, direct DNA amplification or targeted DNA amplification, purification, potentially barcoding multiple samples and subsequently sequencing individually or multiplexed. The purpose of processing is to prepare for subsequently determining the genetic profile. Therefore, by processing, all or part of one or more samples is processed and / or prepared so that it can be delivered to generate a genetic profile. Corresponding treatment measures are described in the prior art (see, for example, RP Schaudies (Ed.): Biological Identification, Woodhead Publishing Series in Electronic and Optical Materials: No. 59, Elsevier 2014, ISBN 978-0-85709-501-5; Jianping Xu: Next-generation Sequencing, Caister Academic Press 2014, ISBN 978-1-908230-33-1; Vijai Bhadauria: Next-generation Sequencing and Bioinformatics for Plant Science, Caister Academic Press 2017, ISBN 978-1-910190-65-4).

[0076] The processing can be performed at the same location where the sample was taken, or alternatively, the sample can be appropriately stored for later processing at a different location. This means that the sample (either immediately or at a later point in time after it has been collected) is transferred for further processing and generation of a genetic profile.

[0077] Analysis of variants present in a gene profile can be performed by direct sequencing of the genomic DNA or RNA region of interest using oligonucleotide probes labeled with appropriate groups; and / or by amplification reactions such as polymerase chain reaction or ligase chain reaction (the products of which can then be analyzed using labeled oligonucleotide probes or several other techniques). In one embodiment, methods for sequencing are described in WO-A 2021 / 228578.

[0078] In one embodiment, sequencing can be performed by using one or more sequencing technologies, including Sanger sequencing, next-generation sequencing, pyrosequencing, nanopore sequencing, GenapSys sequencing, ligation sequencing (SOLID sequencing), single-molecule real-time sequencing, 1on semiconductor sequencing (1on Torrent sequencing), synthesis sequencing (lllumina), combined probe anchor polymerization technology (cPAS-BGI / MGI), nanopore technology, microarray technology, graphene biosensor technology, PCR (polymerase chain reaction) technology, rapid PCR technology and other DNA / RNA amplification technologies such as isothermal amplification such as LAMP (loop-mediated amplification), RPA (recombinase polymerase amplification), nucleic acid sequencing-dependent amplification (NASBA) and transcription-mediated amplification (TMA), as well as epigenetic analysis (e.g., DNA methylation, DNA-protein interaction analysis and chromatin accessibility analysis).

[0079] " amplification " applied to nucleic acid herein is any method that produces one or more copies of nucleic acid, and wherein preferably amplification is exponential growth.A kind of such method for the enzymatic amplification of the specific sequence of DNA is referred to as polymerase chain reaction (PCR), as described in Saiki et al., 1986, Science 230:1350-1354.The primer length used in PCR changes usually in the range of about 10 to 50 or more nucleotides, and typically selects at least about 15 nucleotides to ensure enough specificity.The double-stranded fragment produced is referred to as " amplicon ", and its length can change in the range of 20,000 or more nucleotides to as few as about 30 nucleotides. " genetic marker " used herein is the known variant of the dna sequence dna at specific locus.This variant can be present in individuality due to sudden change or heredity.Genetic marker can be a short dna sequence, such as the sequence (single nucleotide polymorphism, SNP) that changes around a single base pair, or can be a long sequence, such as a minisatellite. Genetic markers can be used to study the relationship between an inherited disease and its genetic cause (eg, a specific mutation in a gene that results in a defective or undesirable form of a protein).

[0080] In addition, the analysis step may include a step of analyzing whether the subject is heterozygous or homozygous for a certain variant. Many different oligonucleotide probe detection formats are known and can be used to implement the present invention. See, for example, US 4,302,204, US 4,358,535, US 4,563,419 and US 4,994,373.

[0081] In one embodiment, analysis may include multiple amplification of DNA (e.g., allele-specific fluorescent PCR). In some embodiments, analysis may include hybridization to a microarray (chip, microbeads, etc.). In some embodiments, analysis may include sequencing the appropriate portion of the gene containing the haploid type to be analyzed. In some embodiments, haploid types may be used to analyze, and the haploid type changes the susceptibility to enzyme digestion by one or more endonucleases. For example, restriction fragment length polymorphism (RFLP) can be used, which refers to the enzyme digestion pattern when multiple restriction endonucleases are applied to DNA. In some embodiments, the presence of one or more haploid types can be determined by allele-specific amplification. In some embodiments, the presence of haploid types can be determined by primer extension. In some embodiments, the presence of haploid types can be determined by oligonucleotide ligation. In some embodiments, the presence of haploid types can be determined by hybridization with a labeled probe.

[0082] Amplification of the selected or targeted nucleic acid sequence in one or more variants can be performed by any suitable method for DNA isolated from a biological sample. Examples of suitable amplification techniques include, but are not limited to, polymerase chain reaction, ligase chain reaction, chain displacement amplification, transcription-based amplification, self-sustaining sequence replication (or "3SR"), nucleic acid sequence-based amplification (or "NASBA"), repair chain reaction (or "RCR"), and boomerang DNA amplification (or "BDA"). Currently preferred is the polymerase chain reaction.

[0083] DNA amplification techniques such as those described above may involve the use of a probe, a pair of probes, or two pairs of probes that, under the same hybridization conditions, bind specifically to DNA encoding a targeted gene but not to DNA encoding a different family member of the targeted gene, and which serve as primers in an amplification reaction for the amplification of the targeted gene DNA or a portion thereof.

[0084] Similarly, one can use a probe, a pair of probes, or two pairs of probes that specifically bind to DNA encoding the variant of interest, but do not bind to other haplotypes under the same hybridization conditions.

[0085] Generally speaking, the oligonucleotide probe used to analyze the DNA encoding the target gene is an oligonucleotide probe that binds to the DNA encoding the haplotype of interest but does not bind to the DNA encoding other haplotypes under the same hybridization conditions. The oligonucleotide probe is labeled with a suitable group, such as those listed below in connection with the antibody.

[0086] Polymerase chain reaction (PCR) can be performed according to known techniques. See, for example, US 4,683,195, 4,683,202, 4,800,159, and 4,965,188. In general, PCR includes first treating a nucleic acid sample with an oligonucleotide primer (which is used for each specific sequence to be analyzed) under hybridization conditions (e.g., in the presence of a thermostable DNA polymerase) so that the extension product of each primer synthesized is complementary to each nucleic acid chain, and the primer is fully complementary to the specific sequence of each chain so as to hybridize therewith, so that the extension product synthesized by each primer, when it is separated from the complementary chain, can serve as a template for the extension product of other primers, and then, if there is one or more sequences to be analyzed, the sample is treated under denaturing conditions so that the primer extension product is separated from its template. These steps are repeated cyclically until the desired degree of amplification is obtained. Analysis of the amplified sequences can be performed by adding to the reaction products an oligonucleotide probe (eg, an oligonucleotide probe of the present invention) capable of hybridizing to the reaction products, the probe carrying a label and then analyzing the label according to known techniques, or by direct visualization on a gel.

[0087] When PCR conditions allow amplification of all alleles of a variant, these species can be distinguished by hybridization with allele-specific probes, restriction enzyme digestion, denaturing gradient gel electrophoresis, or other techniques. Wenham et al. (1991) describe a PCR protocol for determining the genotype of a target gene.

[0088] The term "genotype" refers to the specific allelic form of a gene, which can be defined as the specific nucleotide present at a specific site in a nucleic acid sequence. Genotype can also refer to the paired alleles present at one or more polymorphic loci. For diploid organisms, such as humans, two haploid types constitute the genotype.

[0089] "Genotyping" is any process used to determine the genotype of an individual, for example, by nucleic acid amplification, nucleic acid sequencing, antibody binding, or other chemical analysis. The resulting genotype may be unphased, meaning that it is not known whether the sequence found comes from one parental chromosome or the other.

[0090] In the next step, the results of the genetic analysis of the variants present in the sample are summarized in a genetic profile that lists all variants in the sample and optionally the corresponding sequence information, the organism from which the variant originated and / or the percentage of the variant in the sample.

[0091] In one embodiment, the genetic profile may include one or more of the following variants:

[0092]

[0093]

[0094] Letters represent amino acids according to the International Single-Letter Amino Acid Code (https: / / www.fao.org / 3 / y2775e / y2775e0e.htm). Numbers describe the position of the corresponding amino acid in the gene. The first letter indicates the amino acid present in the wild-type strain or cultivar of the pest; the second letter indicates the genetic variant.

[0095] In another step, one or more biological targets are identified based on the genetic profile.

[0096] The genetic profile provides information on whether one or more pests are present and, if so, which species of pest are present in the sample, and in particular which strain or subtype of the pest. Using publicly available software (such as Basic Local Alignment Search Tool (BLAST; Altschul et al. (1990)), sequence information of variants including genetic profiles can be used to identify corresponding DNA, RNA or protein sequence information of biological targets in sequence databases (e.g., GenBank, Uniprot, Ensembl, etc.). This information can be used to identify one or more corresponding genes or protein sequences encoding one or more biological targets in pests, which can be acted upon to prevent the pests. For example, information about which biological targets in pests can be acted upon and which DNA, RNA or protein sequences are encoded for such biological targets can be stored in a database. It is known that for many pests, which biological targets can be acted upon to control them (see, for example, X. Li et al.: Review on Structures of Pesticide Targets, Int. J. Mol. Sci. 2020, 21, 7144). It is also known which is the biological target for many genetic variants (see the publication FUNGICIDE RESISTANCE ACTION COMMITTEE at https: / / www.frac.info / knowledge-database / downloads, or the INSECTICIDE RESISTANCE ACTION COMMITTEE (https: / / irac-online.org / ) or the HERBICIDE RESISTANCE ACTION COMMITTEE (https: / / www.hracglobal.com / ). Such publicly available information may be compiled and stored in one or more databases.

[0097] In another step, one or more active ingredients that act on one or more biological targets are determined. Just as there is information about the biological targets of pests, there is also information about which biological targets multiple active ingredients act on (see, for example, L.-C. Mei et al.: Pesticide Informatics Platform (PIP): An International Platform for Pesticide Discovery, Residue, and Risk Evaluation, J. Agric. Food Chem. 2022, 70, 6617-6623; publication FUNGICIDE RESISTANCE ACTION COMMITTEE at https: / / www.frac.info / knowledge-database / downloads or INSECTICIDE RESISTANCE ACTION COMMITTEE (https: / / irac-online.org / ) or HERBICIDE RESISTANCE ACTION COMMITTEE (https: / / www.hracglobal.com / )). This information can also be collected and stored in one or more databases so that it can be read therefrom.

[0098] It is also conceivable that the step of "determining one or more biological targets based on the genetic profile" and the step of "determining one or more active ingredients that act on the one or more biological targets" are performed in one step, that is, based on the one or more identified pests, one or more active ingredients are directly identified (for example, obtained from one or more databases), which can be used to control the one or more pests.

[0099] In a further step, the desired efficacy of a crop protection product comprising one or more active ingredients is determined.

[0100] Determination of expected efficacy is accomplished using a model in which, for a large number of crop protection products, their efficacy is correlated with one or more variants and / or one or more genetic profiles of one or more pests. In other words, the model correlates the efficacy of the crop protection product against the pest with the variant or genetic profile of the pest.

[0101] The presence in a field of pest variants that have developed or evolved resistance to a pesticide can have a significant impact on the efficacy of the pesticide.

[0102] "Resistance" refers to an acquired, heritable reduction in the sensitivity of a pest to a specific mode of action. There are two types of resistance: locus-specific resistance (also called target site resistance) and metabolic resistance (also called non-target site resistance). The present invention relates to both types of resistance.

[0103] Since the genetic profile includes information about existing variants, emerging or existing resistance can be inferred from the genetic profile.

[0104] For the genetic analysis of variants present in the genetic spectrum, it should be observed whether such variants are present in the sample, which have been subjected to quantitative or qualitative analysis and indicate that the pest may develop, is developing or has developed resistance to one or more active ingredients and / or crop protection products. In the identification of variants, it can be observed whether there are known variants that will produce resistance in the sample. In addition, in the case of metabolic resistance, the amount of corresponding RNA including one or more variants in the pest can be used as a resistance marker. In one embodiment, the variant identified is neither consistent with a known resistance marker nor consistent with a variant of a non-resistant pest. This new variant may point to a newly formed resistance and / or may indicate a new resistance marker.

[0105] As used herein, "resistance marker" refers to a genetic variant associated with increased resistance to one or more crop protection products. It may also refer to a genetic variant associated with a specific response to one or more crop protection products. By sequencing and analyzing one or more variants as described above, one or more resistance markers can be identified in a sample.

[0106] For a large number of pests, it is known which variants indicate that the pest has developed or is developing resistance to a crop protection product in the target site. Within the meaning of the present disclosure, this type of variant and / or its number is a resistance marker. Several examples are listed below, including examples for the purpose of illustrating the process of discovering variants that represent new resistance markers.

[0107] Acetolactate synthase (ALS or AHAS) is an enzyme involved in the formation of the branched-chain amino acids valine, leucine, and isoleucine in many prokaryotes and eukaryotes. This enzyme is the locus (target) of a series of herbicides known as ALS-inhibiting herbicides: sulfonylureas, imidazolines, triazolopyrimidines, pyrimidylthiobenzoates, and sulfonylaminocarbonyltriazololines. A large number of broadleaf and grass weeds are known to exhibit resistance to ALS-inhibiting herbicides (http: / / www.weedscience.org). J. Rey-Caballero et al. investigated the mechanisms of resistance in corn poppy (Papaver rhoeas) to ALS-inhibiting herbicides (Pesticide Biochemistry and Physiology 138 (2017) 57-65). They found that six amino acids were exchanged in three multiply resistant populations. The gene sequence corresponding to the amino acid is a resistance marker in the sense of the present invention.

[0108] JAC Gardin et al. showed which genetic factors are responsible for the resistance of Alopecurus myosuroides to ALS-inhibiting herbicides (BMC Genomics (2015) 16:590). These can be used as resistance markers in the sense of the present invention.

[0109] There are many studies on gene families encoding cytochrome P450, glutathione S-transferase, glycosyltransferase and ABC transporter proteins in plants (broadleaf / grass weeds) and their involvement in the formation of non-target resistance to herbicides (for example, see JS Yuan et al.: Non-target-site herbicide resistance: a family business, TRENDS in Plant Science, Vol. 12, No. 1 (2006), pp. 6 to 13, and publications cited therein). The disclosed resistance-causing genes and gene families are resistance markers within the meaning of the present invention.

[0110] RH Wrench-Constant summarized the genes and gene families involved in the development of resistance to insecticides in insects (Genetics, Vol. 194 (2013) 807-815). The disclosed genes and gene families that cause resistance are resistance markers in the sense of the present invention.

[0111] The so-called diamide insecticides (http: / / www.alanwood.net / pesticides / class_insecticides.html) are particularly effective control agents against butterflies and moths (Lepidoptera), an example being the tomato spider mite moth.

[0112] In the case of the tomato leafminer (Tuta absoluta), resistance to diamide insecticides has been increasingly observed. E. Roditakis et al. were able to pinpoint specific mutations that are the genetic cause of resistance (Insect Biochemistry and Molecular Biology 80 (2017) 11-20). The DNA sequence affected by the mutation is a resistance marker within the meaning of the present invention.

[0113] Y. Pan et al. showed that genetic factors play a role in the resistance of the aphid Aphis gossypii to the insecticide spirotetramat (Insect Molecular Biology (2017) 26(4), 383-391 doi: 10.1111). These factors can be used as resistance markers in the sense of the present invention.

[0114] Z. Ma et al. were able to show that in the case of the fungus Blumeriella jaapii, resistance to the DMI fungicide is mediated by overexpression of the CYP51 gene (Applied and Environmental Microbiology, April 2006, pp. 2581-2585). Overexpression of the CYP51 gene is a resistance marker in the sense of the present invention.

[0115] S. Omrane et al. showed that in the case of the fungus Zymoseptoria tritici, genetic factors contribute to resistance to a large number of fungicides (mSphere2:e00393-17. https: / / doi.org / 10.1128 / mSphere.00393-17). These factors can be used as resistance markers in the sense of the present invention.

[0116] Methods for identifying resistance markers in the field based on DNA and / or RNA sequencing have been disclosed (WO2011 / 067559A1, WO2012 / 042226A1, WO2013 / 041878A1, WO2013 / 098561A1, WO2013 / 121224A1, WO2014 / 064443A1, WO2015 / 140535A1, WO2015 / 150786A1, WO2015 / 173587A1, WO2016 / 059436A1, WO201 / 6059427A1, WO2017 / 203269A1). Oxford Nanopore Technologies Ltd. provides commercially available corresponding sequencing tools (see, for example, https: / / nanoporetech.com; M. Loose et al.: Real-time selective sequencing using nanopore technology, Nature Method, Vol. 13, No. 9 (2016), 751-758).

[0117] The model may be or include one or more databases storing data on the efficacy of a plurality of crop protection products or a plurality of active ingredients against one or more pests and / or one or more variants of pests.

[0118] Efficacy data for a crop protection product or active ingredient can include EC50 values, mortality, minimum effective rate, disease incidence, disease severity, and / or crop (eg, harvest) loss.

[0119] Efficacy data can be generated in field trials, greenhouse trials, in vitro experiments in the laboratory, or in vivo experiments.

[0120] EC50 values can be measured, for example, in a laboratory. In the laboratory, the efficacy of one or more crop protection products in controlling one or more genetic variants of one or more pests can be measured under controlled conditions. The measured values can be stored in a database. If a specific variant of a pest is found in a sample, the EC50 values of one or more crop protection products associated with that variant can be retrieved from the database. If different variants of a pest are present in the sample, the EC50 values of one or more crop protection products and / or one or more active ingredients for each variant can be retrieved from the database.

[0121] Similarly, where several different pests are included in the sample, the EC50 values of one or more crop protection products for each of these pests may be retrieved from the database.

[0122] If the gene profile includes quantitative information about the amounts of different pests and / or different variants of pests, an effective EC50 value can be calculated from the respective EC50 values, which indicates how effective the crop protection product is against all pests and / or variants present.

[0123] For example, the effective EC50 value can be the sum of the respective EC50 values multiplied by a scaling factor. For example, if a first pest (or a first variant of a pest) is present in a sample at a proportion of X1% (based on the number of individuals of the pest present in the sample) and a second pest (or a second variant of a pest) is present at a proportion of X2% (based on the number of individuals of the pest present in the sample), the EC50 value of the crop protection product for controlling the first pest (or variant) is EC50(1) and the EC50 value of the crop protection product for controlling the second pest (or variant) is EC50(2), then the effective EC50 value can be calculated using the following formula:

[0124] EC50 eff =X1·EC50(1)+X2·EC50 (2)

[0125] The statements in the preceding section apply mutatis mutandis to other quantities that can be used to describe the efficacy of crop protection products, and not only to EC50 values.

[0126] The efficacy of crop protection products in agricultural or horticultural areas can also be determined. For example, the proportion of crops that are damaged or destroyed by one or more harmful organisms (or variants) despite the application of a crop protection product can be determined.

[0127] If the efficacy of a crop protection product is determined based on crop losses caused by the presence of one or more pests (even if the crop protection product is applied), historical crop loss data can also be used to link the genetic profile of the pest or pests to the efficacy of the crop protection product, provided that the genetic profile of the pests present at the time is also available.

[0128] In one embodiment, the determination of the expected efficacy is accomplished using a model in which, for a large number of crop protection products, in addition to their efficacy being related to one or more variants and / or one or more genetic profiles of one or more pests, additional agricultural data is used. This additional agricultural data will allow for a more precise determination of the expected efficacy. Agricultural data may include: (a) information about the time point of sampling; (b) additional information about variants, such as the frequency of occurrence of variants in the pest population, information about whether the variant has been characterized by its effect on the structure or binding properties of the biological target; (c) additional information about the crop protection product, such as application parameters, formulation type, amount of one or more active ingredients, additional ingredients in the formulation such as solvents, emulsifiers; (d) additional information about efficacy, such as efficacy from current or historical experiments or trials in a greenhouse, laboratory or agricultural or horticultural area; additional information for the variant to be analyzed or for the location of (d) efficacy data, such as GPS data; (f) current or historical weather data for one or more agricultural or horticultural areas; (g) environmental data for one or more agricultural or horticultural areas, such as soil data, vegetation data such as vegetation index; (h) additional information about pests, such as species, life cycle stage.

[0129] Of particular interest as agricultural data is the so-called fitness penalty associated with variants. Variants may result in altered expression of the biological target in which the identified variant is present, or in structural changes to the biological target. Consequently, crop protection products acting on the biological target may exhibit altered efficacy. Furthermore, biological targets often have essential or even essential functions in pests. Therefore, variants may also result in altered function of the biological target, which in turn affects the pest's fitness. Fitness is defined as the ability of an organism to survive and reproduce. Therefore, a fitness penalty or fitness cost refers to the effect of a variant on the fitness of an organism in the presence or absence of one or more active ingredients or one or more crop protection products (Hawkins and Fraaije, Annual Review of Phytopathology 2018, Vol. 56, pp. 339-360 (https: / / doi.org / 10.1146 / annurev-phyto-080417-050012)). For example, variants that result in overexpression of a biological target have a cost associated with the allocation of additional resources to that expression, which would not be necessary in the absence of a crop protection product acting on that target. Methods for assessing fitness costs are mutagenesis studies, growth, sporulation, or pathogenicity assays of single isolates of the pest collected in agricultural or horticultural areas, isogenic transformations, or experiments that may include different growth temperatures, growth media, osmotic or oxidative stresses.

[0130] Thus, in one embodiment, a method is disclosed wherein the expected efficacy of a crop protection product comprising one or more active ingredients is determined by using a model in which, for a large number of crop protection products, their efficacy and one or more fitness costs associated with one or more variants are related to one or more genetic profiles of one or more pests.

[0131] The model may also be or include a machine learning model. As used herein, such a "machine learning model" may be understood as a computer-implemented data processing architecture. The machine learning model may receive input data and provide output data based on the input data and parameters of the machine learning model. The machine learning model may learn the relationship between the input data and the output data through training. During training, the parameters of the machine learning model may be adjusted to provide a desired output for a given input.

[0132] The process of training a machine learning model involves providing training data for the machine learning algorithm (i.e., learning algorithm) to learn from. The term "trained machine learning model" refers to the model artifact created by the training process. The training data must include the correct answer, called the target. The learning algorithm finds patterns in the training data that map the input data to the target and outputs a trained machine learning model that captures these patterns.

[0133] During training, training data is input into a machine learning model, and the machine learning model generates an output. The output is compared with a (known) target. The parameters of the machine learning model are modified to reduce the deviation between the output and the (known) target to a (defined) minimum.

[0134] In general, a loss function can be used for training, where the loss function quantifies the deviation between the output and the target. The loss function can be chosen in such a way that it rewards a desired relationship between the output and the target and / or penalizes an undesirable relationship between the output and the target. This relationship can be, for example, similarity or dissimilarity or another relationship.

[0135] If, for example, the output and target are numbers, the loss function can be the difference between these numbers. In this case, a high absolute value of the loss function may mean that the parameters of the model need to be changed significantly.

[0136] In the case of vector-valued outputs, for example, a difference metric between the vectors may be selected, such as root mean square error, cosine distance, a norm of the difference vector such as Euclidean distance, Chebyshev distance, a linear norm of the difference vector, a weighted norm, or any other type of difference metric between two vectors, which may be, for example, a desired output (target) and an actual output.

[0137] In the case of higher dimensional outputs (such as two-dimensional, three-dimensional or higher dimensional outputs), for example, an element-wise difference metric may be used. Alternatively or additionally, the output data may be transformed into a one-dimensional vector before computing the loss.

[0138] The modification of model parameters and the reduction of loss can be performed in optimization methods, such as gradient descent.

[0139] The model of the present disclosure can be trained on training data. For each of a plurality of agricultural or horticultural areas, the training data can include: i) one or more genetic profiles of one or more pests in one or more samples taken in the agricultural or horticultural area as input data, and ii) data describing the efficacy of one or more crop protection products against the one or more pests as target data.

[0140] The term "large number" means at least 10, preferably greater than 100.

[0141] Training a machine learning model can include:

[0142] - Feed input data into the machine learning model;

[0143] - receiving a predicted efficacy of one or more crop protection products from a machine learning model;

[0144] - determining a loss that quantifies the deviation of the predicted efficacy from the target data;

[0145] -Modify the parameters of a machine learning model to minimize the loss.

[0146] The machine learning model can be trained on training data, where variants of the training data are determined at different locations across a large number of agricultural and / or horticultural regions and / or at different points in time during a vegetation period and / or multi-year period. Additional data considered is efficacy data for one or more crop protection products of interest under the conditions at those locations and / or points in time.

[0147] The machine learning model can be or include an artificial neural network. An "artificial neural network" (ANN) is a biologically inspired computing model. An ANN typically includes at least three layers of processing elements: a first layer having input neurons (nodes); a kth layer having at least one output neuron (node); and k-2 inner layers, where k is a natural number greater than 2.

[0148] In such a network, input neurons are used to receive input data. If the input data constitutes or includes an n-dimensional vector (e.g., a feature vector) (where n is an integer equal to or greater than 1), there is usually one input neuron for each component of the vector. Output neurons are used to output at least one value. The processing elements of these layers are interconnected in a predetermined pattern with predetermined connection weights between them. Each network node represents a predetermined calculation of the weighted sum of the inputs from the previous node and represents a nonlinear output function. The combined calculation of the network nodes associates the input with the output. It should be noted that ANN can also use a connection bias b, that is, the output y of a neuron given an input x is calculated as y=act(w x+b), where w represents the weight and act represents the activation function.

[0149] After being trained, the connection weights between processing elements in the ANN contain information about the relationship between input data and output data, which can be used to predict new output data from new input data.

[0150] Training estimates the network weights that allow the network to calculate output values close to the target values. The network weights can be initialized to small random values or to the weights of a previously trained network. Training data input is applied to the network, and an output value is calculated for each training example. The network output value is compared to the target value. A backpropagation algorithm is applied to correct the weight values in a direction that reduces the error between the target output and the calculated output. This process is iterated until the error can no longer be reduced further or until a predetermined prediction accuracy is achieved.

[0151] Cross-validation methods can be used to split the data into a training dataset and a validation dataset. The training dataset can be used in backpropagation training of the network weights. The validation dataset is used to verify that the trained network generalizes to make good predictions. The best set of network weights can be used as the set of network weights that best predicts the output of the training data. Similarly, the number of hidden nodes in the network is varied, and the optimal number of hidden nodes for the network that performs best under the conditions of the used dataset is determined.

[0152] Additional data may be included as input data in the training of the machine learning model, such as weather data, soil data, location information, data about the crops grown and their varieties, images showing the phenotype of the area, information about the crop protection products used (e.g., information about the active ingredients contained in the crop protection products, formulation type), application parameters, etc.

[0153] "Application parameters" means any value defining the application of one or more crop protection products, including application rate, application method, application timing, application machinery for each of the one or more crop protection products.

[0154] When using additional agricultural data, the model learns not only the effect that the presence of one or more variants has on the efficacy of the crop protection product, but also the effect that the additional agricultural data has on the efficacy of the crop protection product.

[0155] In one embodiment, the model is or comprises a simulation model that simulates the effects of one or more crop protection products on one or more pests in an agricultural or horticultural area.

[0156] In one embodiment, the model is or includes a mechanistic efficacy model that quantifies the efficacy of one or more crop protection products on one or more species in a region.

[0157] In one embodiment, the genetic profile data includes information regarding the presence, number, and genotype of variants of more than one pest.

[0158] In one embodiment, the pest is selected from fungal, bacterial or plant species.

[0159] In one embodiment, the efficacy of a crop protection product comprising one or more active ingredients is determined based on the resistance level of a certain variant or a certain genetic profile towards the crop protection product.

[0160] The desired efficacy of the crop protection product can be output, for example displayed on a display of the computer system of the present disclosure, printed via a printing device, stored in a data memory and / or transmitted to a separate computer system.

[0161] Based on the output expected efficacy, a user (e.g., a farmer) can see whether the crop protection product that has output the expected efficacy is a suitable product for controlling pests in an agricultural or horticultural area, or whether the efficacy is too low, for example because the pests present in the agricultural or horticultural area have developed resistance to the crop protection product.

[0162] The user can display the expected efficacy of multiple crop protection products to determine which crop protection product has the highest efficacy.

[0163] The computer system / computer program of the present disclosure may also be configured to determine the efficacy of a plurality of crop protection products and identify the number of crop protection products with the highest efficacy, for example, 1, 2, 3, 4, or 5 or more crop protection products.

[0164] The computer system / computer program of the present disclosure may be configured to output the number of pesticides with the highest efficacy.

[0165] In addition to one or more crop protection products, one or more quantities can also be output which should be applied to effectively control pests and prevent resistance.

[0166] If different pests or different variants of one or more pests have been identified at different locations in an agricultural or horticultural area, the computer system / computer program can be configured to output, for each location, one or more crop protection products that are most effective for controlling the pest or variant at that location.

[0167] For example, a map of an agricultural or horticultural area can be output showing, for different sub-areas of the agricultural or horticultural area, which crop protection products have the highest efficacy against pests or variants in said sub-areas.

[0168] The user may be prompted to select a crop protection product from a list of crop protection products for each sub-region. However, it is also possible that the computer system / computer program is configured to select only one crop protection product (e.g., the product with the highest efficacy) for each sub-region and list it on the map of the agricultural or horticultural region.

[0169] The computer system / computer program is further configured to specify, for a sub-plot, an amount of a crop protection product that should be applied to said sub-plot for controlling pests present in the sub-plot.

[0170] The computer system / computer program may also be configured to output an application map, i.e. a map of an agricultural or horticultural area in which, for a sub-area, the amount of one or more crop protection products to be applied in said sub-area is specified for controlling one or more pests and / or preventing resistance.

[0171] Such an application map can also be transmitted via a network to agricultural or gardening machinery such as a vehicle, drone or robot, which applies one or more crop protection products according to the application map. Such an application map is also called an application plan.

[0172] An "application plan" refers to a script that can execute commands on one or more types of agricultural machinery to cause the agricultural machinery to perform its functions at a specific time and location, optionally using GPS signals for automated control. In one embodiment, the agricultural machinery is a sprayer. Examples of sprayers are self-propelled sprayers, tractor-mounted sprayers, backpack sprayers, and ATV sprayers. In one embodiment, an application plan is used to control a sprayer suitable for spraying one or more crop protection products, so that a specific amount of one or more crop protection products is sprayed at a specific time and / or location using specific application parameters. Those application parameters include: the recommended type of sprayer; additional components for the sprayer, such as boom type and nozzle; the direction of spraying timing and location; the speed of the sprayer; the number of passes during application; the application rate; or the size of the field. Sprayers can be equipped with selective or automatic section control to avoid spraying twice at overlapping locations (e.g., at the end of a field). They can also be equipped with a rate control system, such as pulse width modulation, to ensure consistent spraying. Plot size is an important parameter in the application plan. It is the size of the grid cells and is layered on top of the resistance map or gene profile map to determine whether the crop protection product should be applied in the cell (JC Mayer "Using prescription maps for in field evolution of parameters affecting spraying accuracy of a self-propelled sprayer", Master Thesis North Dacota State University of Agriculture and Applied Science, 2021). The application plan can be machine-readable or human-readable.

[0173] In one embodiment, the time period between sampling of the pests, analysis of the genetic profiles from those samples, generation of the application regimen, and spraying is less than 24 hours, in another embodiment, less than 18, 12, 10, 8, 6, 5, 4, 3, 2, or 1 hour.

[0174] In one embodiment, the time period between sampling of the pests, analysis of the genetic profiles from those samples, generation of the application regimen, and spraying is less than 1 hour.

[0175] In one embodiment, a portable platform is used to generate the genetic profile.

[0176] In one embodiment, the genetic profile is used to predict the occurrence of a pest in an agricultural or horticultural area prior to visualization of the pest in the area.

[0177] The results of the analysis can also be provided in a resistance map. A "resistance map" represents a portion of the earth's surface, wherein for multiple locations on the earth's surface, records are kept of whether pests have been observed at the corresponding locations, where resistance to crop protection products exists, or where known or unknown potential resistance has arisen. The resistance map is preferably a digital reference map. The term "digital" means a map that can be processed by a machine (usually a computer system). "Processing" refers to the known methods of electronic data processing (EDP). Therefore, a "digital resistance map" is a digital representation of a portion of the earth's surface, wherein for multiple locations on the earth's surface, records are kept of whether pests have been observed at the corresponding locations, where resistance to crop protection products exists, or where known or unknown potential resistance has arisen. The digital resistance map is preferably a digital representation of a field, or a digital representation of a field including adjacent areas, or a digital representation of an area. Separate resistance maps can be generated for individual pests, and / or for individual crop protection products or groups of crop protection products that exhibit the same active ingredient or the same chemical / biological type (e.g., chemical structure type) or the same mechanism of action or the same action locus (target site). Separate digital resistance maps can be linked to one another, i.e., virtually superimposed on one another. Furthermore, digital resistance maps can indicate which crops grow in which regions and / or which crop protection products are suitable for controlling pests in said regions and / or which efficacy the corresponding crop protection products have.

[0178] In one embodiment, for each location on the digital resistivity map having one or more analysis results, one or more occasions when the corresponding analysis was performed are recorded.

[0179] In one embodiment, multiple digital resistance maps are linked to each other in a manner to show the development of one or more resistances over time.

[0180] Preferably, the digital resistance map can be combined with other digital maps, for example, digital maps related to soil type, water level, planted crop plants, temperature (at a defined time and / or for a defined time span, for example, in the form of average temperature and / or minimum temperature and / or maximum temperature), precipitation (at a defined time and / or for a defined time span, for example, in the form of average precipitation and / or minimum precipitation and / or maximum precipitation), insolation, air mass movement (wind direction and wind force), past infestations by one or more pests, agricultural measures taken (for example, sowing, watering, tilling, application of crop protection agents, administration of nutrients, etc.), etc. The parameter values recorded in the digital map can be measured values and / or predicted values.

[0181] Operations according to the teachings herein may be performed by at least one computer system specially constructed for the desired purposes or by a general-purpose computer system specially configured for the desired purposes via at least one computer program stored in a typical non-transitory computer-readable storage medium.

[0182] A "computer system" is an electronic data processing system that processes data by means of programmable computational rules. Such a system generally includes a "computer," which includes a processor for performing logical operations, and also includes peripheral devices.

[0183] In computer technology, a "peripheral device" refers to any device connected to a computer that is used to control the computer and / or serves as an input or output device. Examples include monitors (screens), printers, scanners, mice, keyboards, drives, cameras, microphones, speakers, and so on. In computer technology, internal ports and expansion cards are also considered peripherals.

[0184] Computer systems today are generally divided into desktop PCs, portable PCs, laptop computers, notebook computers, netbook computers, and tablet PCs, as well as so-called handheld devices (eg, smartphones); all of these systems can be used to implement the present invention.

[0185] The term "non-transitory" is used herein to exclude transient, propagating signals or waves, but to include any volatile or non-volatile computer storage technology suitable for the present application.

[0186] The term "computer" should be broadly interpreted to cover any type of electronic device with data processing capabilities, including (by way of non-limiting example) personal computers, servers, embedded cores, computing systems, communication devices, processors (e.g., digital signal processors (DSPs)), microcontrollers, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), etc.), and other electronic computing devices.

[0187] The term "processing" as used above is intended to include any type of computation or manipulation or transformation of data represented as physical (e.g., electronic) phenomena that may occur or reside, for example, in registers and / or memory of at least one computer or processor. The term processor includes a single processing unit, or a plurality of distributed or remote such units.

[0188] Figure 1 A computer system (1) according to some example embodiments of the present disclosure is shown in more detail. A computer may include a number of components, of which there may be one or more, such as a processing unit (20) connected to a memory (50) (e.g., a storage device).

[0189] The processing unit (20) may be comprised of one or more processors alone, or one or more processors in combination with one or more memories. A processing unit is generally any computer hardware capable of processing information, such as data, computer programs, and / or other suitable electronic information. A processing unit is comprised of a collection of electronic circuits, some of which may be packaged as an integrated circuit or multiple interconnected integrated circuits (integrated circuits are sometimes more commonly referred to as "chips"). A processing unit may be configured to execute a computer program, which may be stored on the processing unit or in a memory (50) of the same computer or another computer.

[0190] The processing unit (20) can be several processors, multi-core processors or some other types of processors, depending on the specific implementation. In addition, the processing unit can be implemented using several heterogeneous processor systems, in which the main processor and one or more auxiliary processors are present on a single chip. As another illustrative example, the processing unit can be a symmetrical multi-processor system including multiple processors of the same type. In another embodiment, the processing unit can be embodied as or include one or more ASICs, FPGAs, etc. Therefore, although the processing unit can execute a computer program to implement one or more functions, the processing unit of multiple embodiments can implement one or more functions without the help of a computer program. In any case, the processing unit can be appropriately programmed to implement the functions or operations according to the example embodiments of the present disclosure.

[0191] Memory (50) is generally any computer hardware capable of storing information (e.g., data, computer programs (e.g., computer readable program code (60)), and / or other suitable information) in a temporary and / or permanent manner. Memory can include volatile memory and / or non-volatile memory and can be fixed or removable. Examples of suitable memory include random access memory (RAM), read-only memory (ROM), a hard drive, flash memory, a thumb drive, a removable computer disk, an optical disk, magnetic tape, or some combination thereof. An optical disk can include a compact disk - read only memory (CD-ROM), a compact disk - read / write (CD-R / W), a DVD, a Blu-ray disc, etc. In many cases, memory can be referred to as a computer-readable storage medium. A computer-readable storage medium is a non-transitory device capable of storing information and is distinguished from a computer-readable transmission medium (such as an electronic transient signal capable of carrying information from one location to another). As described herein, a computer-readable medium can generally refer to a computer-readable storage medium or a computer-readable transmission medium.

[0192] In addition to the memory (50), the processing unit (20) may also be connected to one or more interfaces for displaying, transmitting and / or receiving information. The interfaces may include one or more communication interfaces and / or one or more user interfaces. The communication interfaces may be configured to transmit information, such as to other computers, networks, databases, etc., and / or receive information, such as from other computers, networks, databases, etc. The communication interfaces may be configured to transmit and / or receive information via physical (wired) and / or wireless communication links. The communication interfaces may include an interface (41) for connecting to a network, such as using technologies such as cellular telephone, Wi-Fi, satellite, cable, digital subscriber line (DSL), fiber optic, etc. In some instances, the communication interfaces may include one or more short-range communication interfaces (42), which are configured to connect devices using short-range communication technologies (such as NFC, RFID, Bluetooth, Bluetooth LE, ZigBee, infrared (e.g., IrDA), etc.

[0193] The user interface may include a display (30). The display may be configured to present or otherwise display information to the user, and suitable examples of displays include liquid crystal displays (LCDs), light emitting diode displays (LEDs), plasma display panels (PDPs), and the like. The user input interface (11) may be wired or wireless and may be configured to receive information from the user into the computer system (1), for example, for processing, storage, and / or display. Suitable examples of user input interfaces include microphones, image or video capture devices, keyboards or keypads, joysticks, touch-sensitive surfaces (separate from or integrated into the touch screen), and the like. In some embodiments, the user interface may include automatic identification and data capture (AIDC) technology (12) for machine-readable information. This may include bar codes, radio frequency identification (RFID), magnetic strips, optical character recognition (OCR), integrated circuit cards (ICC), and the like. The user interface may also include one or more interfaces for communicating with peripheral devices such as printers.

[0194] As described above, program code instructions can be stored in a memory and executed by a processing unit programmed thereby to implement the functions of the systems, subsystems, tools and their corresponding elements described herein. As will be understood, any suitable program code instructions can be loaded onto a computer or other programmable device from a computer-readable storage medium to generate a specific machine so that the specific machine becomes a means for implementing the functions described herein. These program code instructions can also be stored in a computer-readable storage medium, which can instruct a computer, a processing unit or other programmable device to perform its function in a particular manner, thereby generating a specific machine or a specific product. The instructions stored in a computer-readable storage medium can generate products, wherein the products become a means for implementing the functions described herein. Program code instructions can be retrieved from a computer-readable storage medium and loaded into a computer, a processing unit or other programmable device to configure a computer, a processing unit or other programmable device to perform the operations to be performed on a computer, a processing unit or other programmable equipment or by them.

[0195] The retrieval, loading, and execution of the program code instructions may be performed sequentially, such that one instruction is retrieved, loaded, and executed at a time. In some example embodiments, the retrieval, loading, and / or execution may be performed in parallel, such that multiple instructions are retrieved, loaded, and / or executed together. The execution of the program code instructions may produce a computer-implemented process, such that the instructions executed by the computer, processing circuitry, or other programmable device provide operations for implementing the functionality described herein.

[0196] Execution of instructions by the processing unit or storage of instructions in a computer-readable storage medium supports combined operations for implementing the specified functions. In this manner, the computer system (1) may include a processing unit (20) and a computer-readable storage medium or memory (50) connected to the processing circuitry, wherein the processing circuitry is configured to execute computer-readable program code (60) stored in the memory. It will also be understood that one or more functions and combinations of functions may be implemented by a dedicated hardware computer system and / or processing circuitry that performs the specified functions, or by a combination of dedicated hardware and program code instructions.

[0197] Figure 2 One embodiment of a computer-implemented method of the present disclosure is schematically illustrated by way of example in the form of a flow chart.

[0198] The method (100) comprises the following steps:

[0199] (110) receiving a genetic profile of one or more pests from an agricultural or horticultural area, the genetic profile comprising one or more variants;

[0200] (120) identifying one or more biological targets based on the genetic profile;

[0201] (130) identifying one or more active ingredients that act on one or more biological targets;

[0202] (140) using a model to determine the expected efficacy of a crop protection product comprising one or more active ingredients in which, for a large number of crop protection products, their efficacy is associated with one or more genetic profiles of one or more pests;

[0203] (150) Output the desired efficacy of the crop protection product.

[0204] In a non-limiting embodiment, the genetic profile of step (110) is described by the following matrix:

[0205]

[0206]

[0207] In a non-limiting example, the results of determining the biological target in step (120) are described by the following matrix:

[0208]

[0209] In a non-limiting example, the results of determining the one or more active ingredients in step (130) are described by the following list, according to information from FRAC Code List 2022 (https: / / www.frac.info):

[0210] Active ingredients that act on CYP51: tebuconazole, prothioconazole, mefentrifluconazole, difenconazole, metconazole.

[0211] In step (140), the desired efficacy of the crop protection product comprising one or more active ingredients is determined. In a non-limiting embodiment, the desired efficacy result output in step (150) is as follows:

[0212]

[0213]

[0214] Figure 3 The process of training a machine learning model is illustrated by way of example. The machine learning model MLM is trained based on training data TD.

[0215] For each of a plurality of agricultural or horticultural areas, the training data TD includes: i) one or more genetic profiles GP of one or more pests in one or more samples taken in the agricultural or horticultural area; and ii) efficacy data ED of one or more crop protection products against the one or more pests. The training data TD may include additional data FD as described herein.

[0216] One or more genetic profiles GP and optionally further data FD are input into a machine learning model MLM. The machine learning model is configured to generate expected efficacy data eED for one or more crop protection products based at least in part on the input data and model parameters. The expected efficacy data eED are compared with the efficacy data ED. This is done by using a loss function LF, which quantifies the deviation between the expected efficacy data eED and the efficacy data ED. For each pair of expected efficacy data eED and efficacy data ED, a loss value is calculated. During training, the model parameters are modified in such a way that the loss value is reduced to a defined minimum value. The purpose of training is to enable the machine learning model to generate an output that is as close as possible to the corresponding target for each input data. Once the defined minimum value is reached, the (currently fully trained) machine learning model can be used to predict the output for new input data (input data that has not been used during training and the target is generally not (yet) known).

Claims

1. A computer-implemented method comprising: - receiving a genetic profile of one or more pests from an agricultural or horticultural area, said genetic profile comprising one or more variants; - determining one or more biological targets based on the genetic profile; - Identify one or more active ingredients that act on one or more biological targets; - determining the expected efficacy of a crop protection product comprising one or more active ingredients using a model in which, for a large number of crop protection products, their efficacy is correlated with one or more genetic profiles of one or more pests; -Deliver the desired efficacy of crop protection products.

2. The method according to claim 1, further comprising: - selecting at least one crop protection product from one or more crop protection products based on its desired efficacy against one or more pests; - outputting additional application information about at least one selected crop protection product.

3. The method according to claim 1 or 2, further comprising: - generating an application plan comprising one or more crop protection products for use in each agricultural or horticultural area and suitable for application on agricultural machinery.

4. The method according to any one of claims 1 to 3, wherein additional agricultural data is associated with the one or more genetic profiles when determining the desired efficacy.

5. The method according to any one of claims 3 to 4, further comprising one or more of the following steps: - displaying the administration regimen to the user; - performing automatic ordering of one or more crop protection products included in the application regimen; - applying in each agricultural or horticultural area one or more identified crop protection products included in the application scheme; - generating a resistance map based on the application schedule, wherein the resistance map specifies the application of the crop protection products included in the application schedule in the agricultural or horticultural area, and displaying and / or storing the resistance map and / or transmitting the resistance map to a device for applying one or more plant protection products included in the application schedule, in, Any of these steps can occur in any order.

6. The method according to any one of claims 1 to 5, wherein the model is or comprises a trained machine learning model, the machine learning model being trained on training data of one or more agricultural datasets comprising: i) one or more genetic profiles of one or more pests in an agricultural or horticultural area as input data; and ii) efficacy data regarding the efficacy of one or more crop protection products against one or more pests as target data, wherein training the machine learning model comprises: (a) inputting one or more genetic profiles into the machine learning model; (b) receiving predicted efficacy data from the machine learning model; (c) determining a loss that quantifies the deviation between the predicted efficacy data and the efficacy data of the target data; and (d) modifying parameters of the machine learning model to minimize the loss.

7. The method according to any one of claims 1 to 6, wherein the model is or comprises a simulation model that simulates the effect of one or more crop protection products on one or more pests.

8. The method of any one of claims 1 to 6, wherein the model is or comprises a mechanistic efficacy model that quantifies the efficacy of one or more crop protection products against one or more pests.

9. A computer system comprising a processor and a memory storing an application program, wherein the application program is configured to perform an operation when the processor executes the application program, the operation comprising: - receiving a genetic profile of one or more pests from an agricultural or horticultural area, said genetic profile comprising one or more variants; - determining one or more biological targets based on the genetic profile; - Identify one or more active ingredients that act on one or more biological targets; - determining the expected efficacy of a crop protection product comprising one or more active ingredients using a model in which, for a large number of crop protection products, their efficacy is correlated with one or more genetic profiles of one or more pests; -Deliver the desired efficacy of crop protection products.

10. The computer system of claim 9, wherein the operations further comprise: - selecting at least one crop protection product from one or more crop protection products based on its desired efficacy against one or more pests; - outputting additional application information about at least one selected crop protection product.

11. The computer system according to claim 9 or 10, further comprising: - generating an application plan comprising one or more crop protection products for use in each agricultural or horticultural area and suitable for application on agricultural machinery.

12. A computer system according to any one of claims 9 to 11, wherein additional agricultural data is associated with the one or more genetic profiles in determining the expected efficacy.

13. The computer system according to claim 11 or 12, further comprising one or more of the following steps: - displaying the administration regimen to the user; - performing automatic ordering of one or more crop protection products included in the application regimen; - applying one or more crop protection products included in the application scheme in each agricultural or horticultural area; - generating a resistance map based on the application schedule, wherein the resistance map specifies the application of one or more crop protection products included in the application schedule in an agricultural or horticultural area, and displaying and / or storing the resistance map and / or transmitting the resistance map to a device for applying one or more plant protection products included in the application schedule, in, Any of these steps can occur in any order.

14. The method of any one of claims 1 to 8, wherein the model is or comprises a trained machine learning model, the machine learning model being trained on training data of one or more agricultural datasets comprising: i) one or more genetic profiles of one or more pests from an agricultural or horticultural area as input data; and ii) efficacy data regarding the efficacy of one or more crop protection products against one or more pests as target data, wherein training the machine learning model comprises: (a) inputting one or more genetic profiles into the machine learning model; (b) receiving predicted efficacy data from the machine learning model; (c) determining a loss that quantifies the deviation between the predicted efficacy data and the efficacy data of the target data; and (d) modifying parameters of the machine learning model to minimize the loss.

15. The method according to any one of claims 1 to 8, wherein the model is or comprises a simulation model that simulates the effect of one or more crop protection products on one or more pests.

16. The method of any one of claims 1 to 8, wherein the model is or comprises a mechanistic efficacy model that quantifies the efficacy of one or more crop protection products against one or more pests.

17. The method according to any one of claims 9 to 13, wherein the model is or comprises a trained machine learning model, the machine learning model being trained on training data of one or more agricultural datasets comprising: i) one or more genetic profiles of one or more pests from an agricultural or horticultural area as input data; and ii) efficacy data regarding the efficacy of one or more crop protection products against one or more pests as target data, wherein training the machine learning model comprises: (a) inputting one or more genetic profiles into the machine learning model; (b) receiving predicted efficacy data from the machine learning model; (c) determining a loss that quantifies the deviation between the predicted efficacy data and the efficacy data of the target data; and (d) modifying parameters of the machine learning model to minimize the loss.

18. The method according to any one of claims 9 to 13, wherein the model is or comprises a simulation model that simulates the effect of one or more crop protection products on one or more pests.

19. The method of any one of claims 9 to 13, wherein the model is or comprises a mechanistic efficacy model that quantifies the efficacy of one or more crop protection products against one or more pests.

Images