Genome annotation method and electronic device

Abstract

The invention provides a genome annotation method and an electronic device. The genome annotation method comprises the following steps: S1) performing gene structure prediction on a genome by adopting multiple modes to obtain multiple prediction gene sets; s2) performing gene integration on the multiple predictive gene sets by using an EVM tool to obtain an integrated gene set; s3) performing BUSCO evaluation on the integrated gene set to obtain an integrated gene set evaluation file; s4) performing BUSCO evaluation on the genome to obtain a genome evaluation file; and S5) correcting the integrated gene set evaluation file by using the genome evaluation file to obtain a corrected gene set, wherein the gene structure prediction comprises transcriptome prediction, de novo prediction and homologous prediction.The genome annotation method can significantly improve the accuracy and integrity of gene annotation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of genomics and bioinformatics, and particularly relates to a genome annotation method and an electronic device. BACKGROUND

[0002] Genome assembly annotation is a core content in the field of bioinformatics. Genome assembly is a process of generating a genome base sequence by sequence splicing of fragments reads generated by sequencing. The process includes: constructing contigs according to the overlap between sequencing sequences, and using allhic and other software to cluster, sort and adjust the direction of the contigs to obtain a chromosome-level genome. The format is fasta format, including a description line (starting with "> ", containing sequence id or name and other unique information) and a sequence line (nucleic acid sequence represented by ATCG).

[0003] After the genome assembly is completed, annotation is needed, mainly including repeat sequence annotation, non-coding RNA annotation (ncRNA), gene structure annotation and gene function annotation. Before annotation, a gene model needs to be constructed. In the prior art, constructing a gene model mainly includes three strategies: 1) homology-based prediction: using the existing high-quality annotation information of a close relative species to determine the exon boundary and the splicing site through sequence alignment, which is suitable for genes with high conservation; 2) transcriptome-based prediction: using the RNA-seq data of a species to assist in annotation, which can accurately determine the exon region and the splicing site; 3) de novo prediction: predicting the gene structure through an existing probability model.

[0004] Since each of the above methods has advantages and disadvantages, the EvidenceModeler (EVM) tool is usually used to integrate the three annotation results to form a complete gene structure, and subsequent functional annotation is based on this. The annotation result is generally a gff format file (containing 9 columns of information, including sequence id, source, type, position, etc.), and a fasta format cds (coding sequence) file and a protein pep (protein) file are generated based on this.

[0005] However, the gene set obtained by the above integration method has the problem of insufficient integrity in BUSCO evaluation, and the main reason is that eukaryotic genes are divided into exons and introns, intron splicing and exon splicing occur during transcription to form different transcripts. At the same time, a large number of non-coding regions in the genome seriously interfere with gene positioning. This significantly increases the difficulty of gene structure annotation, resulting in missing part of the annotated gene set, lower protein value, and much lower BUSCO value than the genome itself (the genome contains all genetic information, and the BUSCO evaluation result is usually higher). The insufficient BUSCO integrity has become a key bottleneck restricting the accuracy of genome structure annotation. SUMMARY

[0006] The main purpose of the present application is to provide a genome annotation method and electronic device to solve the problem of incomplete genes obtained by the genome annotation in the prior art.

[0007] In order to achieve the above-mentioned purpose, according to the first aspect of the present application, a genome annotation method is provided, which comprises: S1) predicting gene structure of the genome in multiple ways to obtain multiple predicted gene sets; S2) integrating the multiple predicted gene sets by using an EVM tool to obtain an integrated gene set; S3) performing BUSCO evaluation on the integrated gene set to obtain an integrated gene set evaluation file; S4) performing BUSCO evaluation on the genome to obtain a genome evaluation file; S5) correcting the integrated gene set evaluation file by using the genome evaluation file to obtain a corrected gene set; The gene structure prediction includes transcriptome prediction, de novo prediction and homology prediction.

[0008] Further, the correction comprises: aligning the genome evaluation file and the integrated gene set evaluation file to obtain a gene to be backfilled; integrating the gene to be backfilled into the integrated gene set evaluation file; wherein the gene to be backfilled is a gene present in the genome evaluation file but not present in the integrated gene set evaluation file.

[0009] Further, aligning the genome evaluation file and the integrated gene set evaluation file to obtain the gene to be backfilled also obtains a first gene and a second gene; the correction further comprises: comparing the sequence lengths of the first gene and the second gene; when the sequence length of the first gene is longer than the sequence length of the second gene, the first gene is integrated into the integrated gene set evaluation file and the second gene in the integrated gene set evaluation file is removed; when the sequence length of the second gene is longer than the sequence length of the first gene, the second gene in the integrated gene set is retained; wherein the first gene and the second gene have an intersection in the exon region; the first gene is derived from the genome evaluation file; and the second gene is derived from the integrated gene set evaluation file.

[0010] Further, the S2) comprises: gene integration is performed on the plurality of predicted gene sets by using an EVM tool to obtain a preliminary integrated gene set; the preliminary integrated gene set is updated by using a Pasa software to obtain the integrated gene set.

[0011] Further, the software for transcriptome prediction comprises Hisat2 software, Stringtie software, Trinity software and Pasa software; the software for de novo prediction comprises one or more of August software, Snap software or GeneMark; the software for homology prediction is selected from genewise software.

[0012] In order to achieve the above-mentioned purpose, according to the second aspect of the present application, an electronic device for genome annotation is provided, which comprises a predicted gene set acquisition unit, an integrated gene set unit, a BUSCO evaluation unit and a correction unit; wherein the predicted gene set acquisition unit is configured to perform gene structure prediction on a genome in multiple ways to obtain a plurality of predicted gene sets; the integrated gene set unit is configured to perform gene integration on the plurality of predicted gene sets by using an EVM tool to obtain an integrated gene set; the BUSCO evaluation unit comprises a first BUSCO evaluation module and a second BUSCO evaluation module; the first BUSCO evaluation module is configured to perform BUSCO evaluation on the integrated gene set to obtain an integrated gene set evaluation file; the second BUSCO evaluation module is configured to perform BUSCO evaluation on the genome to obtain a genome evaluation file; the correction unit is configured to correct the integrated gene set evaluation file by using the genome evaluation file to obtain a corrected gene set; The gene structure prediction includes transcriptome prediction, de novo prediction and homology prediction.

[0013] Further, the correction unit includes a first correction module; The first correction module includes a first alignment unit and a first backfill unit. The first alignment unit is configured to align the genome evaluation file and the integrated gene set evaluation file to obtain the genes to be backfilled. The first backfill unit is configured to integrate the genes to be backfilled into the integrated gene set evaluation file. The genes to be backfilled are genes that exist in the genome evaluation file but do not exist in the integrated gene set evaluation file.

[0014] Further, the correction unit further includes a second correction module; The second correction module includes a second alignment unit, a comparison unit, an integration unit, a removal unit and a retention unit. The second alignment unit is configured to align the genome evaluation file and the integrated gene set evaluation file to obtain the genes to be backfilled, as well as the first gene and the second gene. The comparison unit is configured to compare the sequence lengths of the first gene and the second gene. The integration unit is configured to integrate the first gene into the integrated gene set evaluation file when the sequence length of the first gene is longer than the sequence length of the second gene. The removal unit is configured to remove the second gene in the integrated gene set evaluation file when the sequence length of the first gene is longer than the sequence length of the second gene. The retention unit is configured to retain the second gene in the integrated gene set when the sequence length of the second gene is longer than the sequence length of the first gene. The first gene and the second gene have an intersection in the exon region. The first gene is derived from the genome evaluation file, and the second gene is derived from the integrated gene set evaluation file.

[0015] Further, the integrated gene set unit includes an integration module and a correction module. The integration module is configured to integrate genes in the plurality of predicted gene sets using an EVM tool to obtain a preliminary integrated gene set. The correction module is configured to update the preliminary integrated gene set using Pasa software to obtain the integrated gene set.

[0016] Further, the transcriptome prediction software includes Hisat2 software, Stringtie software, Trinity software and Pasa software. The de novo prediction software includes one or more of August software, Snap software or GeneMark. The homology prediction software is selected from genewise software.

[0017] According to a third aspect of the present application, a computer readable storage medium is provided, the storage medium comprising a stored program, wherein the program, when executed, controls the method of genome annotation.

[0018] According to a fourth aspect of the present application, a processor is provided, the processor being configured to execute a program, wherein the program, when executed, performs the method of genome annotation.

[0019] According to the technical solution of the present application, the genome is subjected to transcriptome prediction, de novo prediction and homology prediction, and the corresponding predicted gene sets are obtained, then the predicted gene sets are integrated to obtain an integrated gene set. The integrated gene set and the genome are subjected to BUSCO evaluation, the integrated gene set evaluation file is modified by using the BUSCO evaluation file of the genome, and a modified gene set with improved integrity can be obtained, and the genome annotation is performed by using the modified gene set, which can further improve the accuracy of the genome annotation result. BRIEF DESCRIPTION OF DRAWINGS

[0020] The drawings constituting a part of the specification of the present application are used to provide further understanding of the present application, the illustrative embodiments of the present application and the description thereof are used to explain the present application, and do not constitute improper limitation on the present application. In the drawings: Figure 1 A flow chart of genome annotation according to an embodiment of the present application is shown.

[0021] Figure 2 A hardware structure block diagram of genome annotation according to an embodiment of the present application is shown.

[0022] Figure 3 A schematic diagram of an electronic device for genome annotation according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0023] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below in combination with the embodiments.

[0024] TERMS EXPLANATION: Homology-based prediction: Utilize existing high-quality annotation information from closely related species to determine exon boundaries and splice sites through sequence alignment, suitable for genes with high conservation.

[0025] Transcriptome-based prediction: With the aid of RNA-seq data of the species, it can accurately determine the exon region and splice site; use RNA-seq data to infer the transcription of genes in the genome, and then assist in predicting gene structure. This process usually includes using software to read transcriptome data such as Hisat2, Trinity and StringTie, then aligning with genome sequences, and annotating through PASA or other tools.

[0026] De novo prediction: Predict gene structure, including exon and intron boundaries, through existing probability models (e.g., Hidden Markov Model), without reference sequence information. BUSCO (Benchmarking Universal Single-Copy Orthologs) is a key tool for evaluating the completeness of genome assembly and annotation, by aligning with single-copy ortholog databases, with complete (the higher the index, the better), Fragmented (the lower the index, the better), Missing (the lower the index, the better) three indicators to measure quality.

[0027] EVM (Evidence Modeler): A software tool for integrating multiple evidence sources to predict and annotate gene structures in eukaryotic genomes. Its core function is to integrate and evaluate results from different prediction methods (e.g., de novo prediction, homology prediction or transcriptome prediction), by comparing and evaluating the consistency of predictions from different evidence sources, to generate a more reliable and accurate gene annotation. It uses external evidence (such as known protein sequences, transcript sequences, etc.) to correct errors in de novo prediction results, thereby improving the accuracy of gene prediction.

[0028] PASA (Program to Assemble Spliced Alignments): A bioinformatics tool for transcript annotation and gene structure updating, designed specifically for integrating transcriptome sequencing data (RNA-seq) to improve gene annotation.

[0029] The main functions of PASA include: 1) processing transcriptome data: PASA can receive transcriptome data generated by software such as Hisat2, Trinity and StringTie, process and assemble transcripts; 2) alignment with genome sequence: PASA aligns the processed transcript data with the genome sequence, identifies the exon and intron structure in the transcript, and the isomerism between transcripts; 3) annotation update: by comparing the structure of the transcript and the existing gene annotation, PASA can identify and correct errors in the gene model, such as incorrect intron boundaries, missing exons, etc., it can also identify new genes and variable splicing patterns of transcripts; 4) backfilling and expansion: PASA can backfill or expand the gene model based on transcriptome evidence to ensure the integrity and accuracy of the annotation results; 5) redundancy removal and optimization: PASA also includes the function of removing redundant transcripts, and optimizing transcripts and gene models based on length and consistency.

[0030] As mentioned in the background, after the transcriptome prediction, de novo prediction and homology prediction of the genome in the prior art and the integration of the predicted gene set, the integrated gene set obtained has the problem of insufficient integrity in BUSCO evaluation. In order to reduce the gap between the gene set busco value and the genome busco value and supplement the missing proteins, the inventors of the present application attempt to use the high busco evaluation result of the genome itself as a supplementary information source, breaking the limitation of relying only on the integration of multiple annotation methods in the traditional annotation process, and thus the protection scheme of the present application is proposed.

[0031] In a first typical embodiment of the present application, a genome annotation method is provided, which comprises: S1) predicting gene structure of the genome in multiple ways to obtain multiple predicted gene sets; S2) integrating the multiple predicted gene sets by using the EVM tool to obtain an integrated gene set; S3) performing BUSCO evaluation on the integrated gene set to obtain an integrated gene set evaluation file; S4) performing BUSCO evaluation on the genome to obtain a genome evaluation file; S5) using the genome evaluation file to correct the integrated gene set evaluation file to obtain a corrected gene set; The gene structure prediction includes transcriptome prediction, de novo prediction and homology prediction.

[0032] The above genome annotation method can significantly improve the accuracy and integrity of gene annotation. By comprehensively using different prediction methods, the limitations of single prediction method can be made up, the integration ability of EVM further optimizes the gene model, and the BUSCO evaluation provides an objective annotation quality index. Finally, the correction step based on the genome evaluation can supplement the missing genes, improve the BUSCO index of the annotation set, and ensure that the annotation result is closest to the true situation of the genome.

[0033] The annotation method of the present application has clear steps, and the annotation software, evm tool and busco evaluation tool used are all common tools in the industry, without the need to develop new software or equipment, easy to apply on a conventional bioinformatics analysis platform, and can be seamlessly connected with the existing annotation process.

[0034] It should be noted that when BUSCO evaluation is performed on the genome and the integrated gene set, a database of the same BUSCO species needs to be selected for evaluation. By using a unified database, BUSCO genes present in the genome but not present in the integrated gene set can be determined.

[0035] In a preferred embodiment of the present application, the correction includes: comparing the genome evaluation file and the integrated gene set evaluation file to obtain the genes to be supplemented; and integrating the genes to be supplemented into the integrated gene set evaluation file; wherein the genes to be supplemented are genes present in the genome evaluation file but not present in the integrated gene set evaluation file.

[0036] The above correction helps to improve the integrity of the integrated gene set and ensures that all orthologous genes present in the genome are accurately annotated.

[0037] In a preferred embodiment of the present application, the comparison of the genome evaluation file and the integrated gene set evaluation file to obtain the genes to be supplemented also obtains a first gene and a second gene; the correction further includes: comparing the sequence lengths of the first gene and the second gene; when the sequence length of the first gene is longer than the sequence length of the second gene, integrating the first gene into the integrated gene set evaluation file and removing the second gene in the integrated gene set evaluation file; when the sequence length of the second gene is longer than the sequence length of the first gene, retaining the second gene in the integrated gene set; wherein the first gene and the second gene have an intersection in the exon region; the first gene is derived from the genome evaluation file; and the second gene is derived from the integrated gene set evaluation file.

[0038] By comparing the sequence length of the first gene (from the genome evaluation file) with the second gene (from the integrated gene set evaluation file), the longer or more functionally explicit gene is selected for integration, effectively avoiding redundancy in the gene set. This processing ensures the refinement and reliability of the gene set, improving its quality and providing high-quality gene data for subsequent genomics research.

[0039] In a preferred embodiment of the present application, S2) comprises: Integrating the above-mentioned multiple predicted gene sets using EVM tools to obtain a preliminary integrated gene set; Updating the above-mentioned preliminary integrated gene set using Pasa software to obtain the integrated gene set.

[0040] The integration of EVM ensures the diversity and coverage of gene models, while the update of Pasa further improves the accuracy of gene models, especially for dynamic changes and complex splicing patterns of gene structure. This process optimizes the prediction of gene structure, reduces prediction errors, and improves the reliability of annotation.

[0041] In a preferred embodiment of the present application, the software for transcriptome prediction includes Hisat2 software, Stringtie software, Trinity software and Pasa software; The software for de novo prediction includes one or more of August software, Snap software or GeneMark; The software for homologous prediction is selected from genewise software.

[0042] By selecting industry standard software (such as Hisat2, Stringtie, Trinity, Pasa, Genewise, August, Snap, GeneMark, etc.), the consistency and high quality of the prediction results are ensured. The flexible combination of these software can provide the best annotation effect for different species and data types.

[0043] In a second typical embodiment of the present application, an electronic device for genome annotation is provided (see Figure 3 ), which comprises a predicted gene set acquisition unit 10, an integrated gene set unit 20, a BUSCO evaluation unit 30 and a correction unit 40; The predicted gene set acquisition unit 10 is configured to predict gene structure of the genome in multiple ways to obtain multiple predicted gene sets; The integrated gene set unit 20 is configured to integrate the above-mentioned multiple predicted gene sets using EVM tools to obtain an integrated gene set; The BUSCO evaluation unit 30 includes a first BUSCO evaluation module 301 and a second BUSCO evaluation module 302. The first BUSCO evaluation module 301 is configured to perform BUSCO evaluation on the integrated gene set to obtain an integrated gene set evaluation file. The second BUSCO evaluation module 302 is configured to perform BUSCO evaluation on the genome to obtain a genome evaluation file. The correction unit 40 is configured to correct the integrated gene set evaluation file by using the genome evaluation file to obtain a corrected gene set. The gene structure prediction includes transcriptome prediction, de novo prediction, and homology prediction.

[0044] The automatic design of the electronic device enables the annotation method to be efficiently and accurately performed, reduces errors caused by manual intervention, and speeds up the annotation process. The modular structure not only facilitates software updates and maintenance, but also improves the portability and applicability of the method.

[0045] In a preferred embodiment of the present application, the correction unit 40 includes a first correction module 401. The first correction module 401 includes a first alignment unit 4011 and a first backfill unit 4012. The first alignment unit 4011 is configured to align the genome evaluation file and the integrated gene set evaluation file to obtain a gene to be backfilled. The first backfill unit 4012 is configured to integrate the gene to be backfilled into the integrated gene set evaluation file. The gene to be backfilled is a gene that exists in the genome evaluation file but does not exist in the integrated gene set evaluation file.

[0046] In a preferred embodiment of the present application, the correction unit 40 further includes a second correction module 402. The second correction module 402 includes a second alignment unit 4021, a comparison unit 4022, an integration unit 4023, a removal unit 4024, and a retention unit 4025. The second alignment unit 4021 is configured to align the genome evaluation file and the integrated gene set evaluation file to obtain the gene to be backfilled, as well as a first gene and a second gene. The comparison unit 4022 is configured to compare the sequence lengths of the first gene and the second gene. The integration unit 4023 is configured to integrate the first gene into the integrated gene set evaluation file when the sequence length of the first gene is longer than the sequence length of the second gene. The removal unit 4024 is configured to remove the second gene in the integrated gene set evaluation file when the sequence length of the first gene is longer than the sequence length of the second gene. The retention unit 4025 is configured to retain the second gene in the integrated gene set when the sequence length of the second gene is longer than the sequence length of the first gene. The first gene and the second gene have an intersection in an exon region. The first gene is derived from the genome evaluation file, and the second gene is derived from the integrated gene set evaluation file.

[0047] In a preferred embodiment of the present application, the integrated gene set unit 20 comprises an integration module 201 and a correction module 202. The integration module 201 is configured to integrate genes in the plurality of predicted gene sets by using an EVM tool to obtain a preliminary integrated gene set. The correction module 202 is configured to update the preliminary integrated gene set by using a Pasa software to obtain the integrated gene set.

[0048] In a preferred embodiment of the present application, the software for transcriptome prediction comprises Hisat2 software, Stringtie software, Trinity software and Pasa software. The software for de novo prediction comprises one or more of August software, Snap software or GeneMark. The software for homology prediction is selected from genewise software.

[0049] In a third typical embodiment of the present application, a computer readable storage medium is provided, which comprises a stored program, wherein the program controls the method for genome annotation when the program is executed.

[0050] In a fourth typical embodiment of the present application, a processor is provided for executing a program, wherein the program executes the method for genome annotation (see Figure 2 ).

[0051] The method provided in the present application can be executed in a terminal, a computer terminal or a similar computing device. Taking the execution on the terminal as an example, Figure 2 is a hardware structure block diagram of a genome annotation method of an embodiment of the present application. As Figure 2As shown, a terminal may include one or more ( Figure 2 Only one is shown in the diagram. A processor A1 (processor A1 may include, but is not limited to, a microprocessor (MCU) or a programmable logic device (FPGA), etc.) and a memory B1 for storing data are also shown. Optionally, the terminal may further include a transmission device C1 for communication functions and an input / output device D1. Those skilled in the art will understand that... Figure 2 The structure shown is for illustrative purposes only and does not limit the structure of the terminal described above. For example, the terminal may also include components that are more... Figure 2 The more or fewer components shown, or having the same Figure 2 The different configurations shown.

[0052] Memory B1 can be used to store computer programs, such as application software programs and modules, like the computer programs corresponding to the gene structure annotation, gene integration, Busco evaluation, and correction methods in this embodiment of the invention. Processor A1 executes various functional applications and data processing by running the computer programs stored in memory B1, thereby implementing the methods described above. Memory B1 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, memory B1 may further include memory remotely located relative to processor A1, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0053] Transmission device C1 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the terminal's communication provider. In one example, transmission device C1 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, transmission device C1 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0054] Obviously, those skilled in the art should understand that some modules or steps of this application described above can be implemented in general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device, or fabricating them separately as individual integrated circuit modules, or fabricating multiple modules or steps into a single integrated circuit module. Thus, this application is not limited to any particular hardware and software combination.

[0055] The present application realizes hardware and programming of the annotation process by solidifying the annotation method in the storage medium and the processor, and guarantees the stability and speed of the annotation process. This not only facilitates the annotation of large-scale genomic data, but also provides a technical basis for remote data processing and genomic research in a cloud computing environment.

[0056] The present application will be further described in detail below in combination with specific embodiments, which cannot be understood as limiting the scope of the present application.

[0057] Embodiment 1 The present embodiment provides a genomic annotation method for effectively improving the busco index (see Figure 1 ), and the method is as follows: Step 1: Separate annotation: using the denovo annotation method, the annotation method based on homologous species and the annotation method combined with transcriptome data, the target genome is independently annotated to obtain three annotation results; Step 2: evm integration: the three annotation results obtained in step 1 are integrated by the evm tool to obtain a preliminary gene set; Step 3: Pasa update: the integrated gene set obtained in step 2 is extended and corrected to obtain a more complete gene set; Step 4: Genomic busco evaluation: the target genome is evaluated by busco to obtain the information of the busco involved in the genome, which is recorded as the genomic busco list; this step is one of the core innovations of the present application, which provides a high-quality data source for the subsequent supplement step by directly extracting the busco gene set with high integrity from the genome; Step 5: Gene set busco evaluation: the complete gene set obtained in step 3 is evaluated by busco to obtain the information of the busco involved in the gene set, which is marked as the gene set busco list; Step 6: Information merging: by comparing the busco list information in steps 4 and 5, the complete busco in the genome and the missing information in the gene set are screened out and filled into the gene set, and this step realizes the transmission of information from the genome to the annotated gene set, which is a key operation to improve the busco integrity; Step 7: Removing overlapping genes: the merged gene set obtained in step 6 is analyzed to remove the overlapping genes in the integrated gene set to obtain the final gene set. The busco integrity of the gene set is significantly improved. Through the de-overlapping processing, the gain of the supplemented genes is retained, and the gene redundancy is avoided, which guarantees the accuracy of the gene set.

[0058] In step 1, the denovo annotation method adopts gene prediction software such as augustus and geneid; the annotation method based on homologous species is realized by aligning the known gene sequences of homologous species with the target genome; and the annotation method combined with transcriptome data is realized by aligning the unigene sequences obtained by assembling the transcriptome sequencing data with the target genome.

[0059] In step 4, when the busco evaluation is performed on the target genome, the busco database corresponding to the species to which the target genome belongs is selected. The targeted database selection ensures the adaptability of the extracted genome busco gene set to the target species, and improves the supplement effect.

[0060] In step 7, whether the genes overlap is judged by aligning the position information of the genes, and when there is an intersection between the exon regions of two genes, the genes are determined to be overlapping genes. For overlapping genes, the gene with longer length or more explicit functional annotation is retained. This de-overlapping strategy takes into account the structural integrity and functional significance of the genes, and is an important link to ensure the quality of the final gene set.

[0061] Compared with the prior art D1 (CN116758995A) and D2 (CN118398090A), the busco indicators of the method of the present application are significantly improved through multiple project practice verification, and the experimental results are shown in Table 1.

[0062] Table 1 Species Genome busco Conventional annotation busco indicator D1 busco indicator D2 busco indicator Invention busco indicator Hordeum vulgare 99.1% 91.5% 97.4% 97.6% 98.6% Mactra antiquata 96.8% 92.4% 95.5% 95.3% 96.2% Larimichthys crocea 97.5% 93.6% 95.9% 95.7% 97.1% Pyrus spp. 98.7% 92.3% 96.2% 96.8% 98.3% In addition, compared with D1 (CN116758995A) and D2 (CN118398090A), the present application also has significant innovation and superiority in technical path, information utilization and effect realization, as follows: 1) Technical path D1 focuses on parameter optimization in the annotation process, by introducing BUSCO protein library in homologous prediction stage to obtain the first homologous prediction open reading frame, and giving it a high weight of 30-35% in EVM integration. The essence is to improve the BUSCO index by adjusting the parameters in the prediction and integration stages, which still belongs to the category of "step-by-step optimization in the annotation process", and still depends on the intermediate products in the annotation process.

[0063] D2 performs BUSCO evaluation on the gene sets and integration results of transcriptome prediction, de novo prediction and homologous prediction, respectively, and based on the evaluation file, the missing genes are supplemented and the redundant genes are deleted in multiple steps. The core is to use the BUSCO information of the gene sets in each stage to modify the integration results in the annotation process, which still belongs to the logic of "dynamic modification in the annotation process", and needs to rely on the alignment of multiple rounds of predicted gene sets, and the process is more complex and the supplement improvement effect is still based on the traditional annotation method.

[0064] After completing the complete annotation process of EVM integration and Pasa update, the application directly extracts the BUSCO information of the complete and gene set missing in the target genome for supplementation, forming a new path of "annotation completion → genome high BUSCO information directional recovery", without intervention in the prediction parameters or integration weight in the annotation process, and the technical logic is more direct and efficient.

[0065] 2) Information utilization The BUSCO information source of D1 is limited to the prediction results of the BUSCO protein library, which needs to be indirectly obtained through homologous alignment, and is limited by the species adaptability of the protein library, which may affect the supplementation effect due to the difference in sequence conservation in the library.

[0066] D2 depends on the BUSCO evaluation file of each predicted gene set (transcriptome, de novo, and homologous prediction), and its information source is still the intermediate gene set generated during the annotation process, which has not broken the closed loop of "annotation product → annotation product correction", and it is difficult to use the original genetic information in the genome that is not captured by the annotation process.

[0067] The application innovatively uses the high BUSCO evaluation results of the target genome itself as the core supplement source, as the genome carries the complete genetic information of the species, and its BUSCO value is naturally higher than that of the annotated gene set. Therefore, the supplementation can directly recover the missing key genes in the annotation process, avoiding the dependence on external protein libraries or intermediate products, and the information utilization is more direct and comprehensive.

[0068] 3) More significant effect and more efficient operation Through experimental verification, the BUSCO completeness of the annotated gene sets of Hordeum vulgare, Mactra antiquata, Larimichthys crocea, and Pyrus spp. is improved by 7.1%, 3.8%, 3.5%, and 6%, respectively. D1 is improved by 5.9%, 3.1%, 2.3%, and 3.9%, respectively. D2 is improved by 6.1%, 2.9%, 2.1%, and 4.5%, respectively. The improvement effect of the application is better (see Table 1).

[0069] D1 needs to finely adjust the weight of multiple predicted open reading frames during EVM integration (such as 6 reading frame weight grading settings), and D2 needs to perform multiple rounds of BUSCO evaluation and alignment on 4 predicted gene sets, both of which involve complex parameter adjustment and calculation alignment, consuming a large amount of computing power and time.

[0070] The application only needs to compare and deoverlap the BUSCO lists of the genome and the gene set after the annotation is completed, without the need for process parameter optimization, significantly saving computing resources and operation time, and meeting the requirement of efficiency.

[0071] In summary, the application uses the new technical path of "annotated genome direct retrieval", fully utilizes the complete genetic information of the genome, and is superior to D1 and D2 in improving the effect, operation efficiency and resource consumption, and provides a more innovative and practical solution for improving the BUSCO integrity of genome annotation.

[0072] From the above description, it can be seen that the above-mentioned embodiments of the application achieve the following technical effects: the application supplements the gene set involved in busco in the genome to the preliminary gene set obtained by evm integration, and removes the overlapping genes, fully utilizes the characteristics of high busco evaluation result of the genome, effectively makes up for the deficiency of the preliminary gene set in busco integrity, and significantly improves the busco integrity of the preliminary gene set. Experimental data show that the busco integrity of the gene set obtained by the method of the application is improved by more than 3-7% compared with the preliminary gene set, which is much higher than the optimization effect of the prior art.

[0073] Finally, it should be noted that the BUSCO index reflects the quality of genome structure annotation. The conserved genes only account for 1%-20% of the genome, and most of them have been captured by the existing annotation process. When the BUSCO integrity reaches more than 90%, the remaining unannotated genes are mostly high-complexity region genes (such as located in a region with high repetitive sequences, abnormal GC content or low chromatin openness, conventional prediction tools are difficult to detect due to ambiguous sequence characteristics), low-expression genes (with very low coverage in transcriptome data, it is difficult to verify the transcript support by PASA and other tools) and species-specific conserved genes (with large differences in sequence from the reference species in the BUSCO database, homologous alignment is difficult to match). The lack of annotation of these genes cannot be solved by simple parameter optimization, and it is necessary to break through the limitations of the existing annotation logic. Therefore, the improvement of the BUSCO integrity of the gene set needs to overcome multiple technical obstacles, and the improvement of 3%-7% is a significant breakthrough.

[0074] The above only describes the preferred embodiments of the application and is not intended to limit the application. For those skilled in the art, the application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the application shall be included in the protection scope of the application.

Claims

1. A method of genome annotation, characterized by, The method comprises: S1) predicting gene structures of a genome in multiple ways to obtain multiple predicted gene sets; S2) integrating genes of the multiple predicted gene sets by using an EVM tool to obtain an integrated gene set; S3) performing BUSCO evaluation on the integrated gene set to obtain an integrated gene set evaluation file; S4) performing BUSCO evaluation on the genome to obtain a genome evaluation file; S5) correcting the integrated gene set evaluation file by using the genome evaluation file to obtain a corrected gene set; The gene structure prediction comprises transcriptome prediction, de novo prediction, and homologous prediction.

2. The method of claim 1, wherein, The correction comprises: aligning the genome evaluation file and the integrated gene set evaluation file to obtain genes to be backfilled; integrating the genes to be backfilled into the integrated gene set evaluation file; wherein the genes to be backfilled are genes that exist in the genome evaluation file but do not exist in the integrated gene set evaluation file.

3. The method of claim 2, wherein, The alignment of the genome evaluation file and the integrated gene set evaluation file obtains the genes to be backfilled, and also obtains a first gene and a second gene; the correction further comprises: comparing sequence lengths of the first gene and the second gene; when the sequence length of the first gene is longer than the sequence length of the second gene, integrating the first gene into the integrated gene set evaluation file and removing the second gene in the integrated gene set evaluation file; when the sequence length of the second gene is longer than the sequence length of the first gene, retaining the second gene in the integrated gene set; wherein the first gene and the second gene have an intersection in an exon region; the first gene is derived from the genome evaluation file; and the second gene is derived from the integrated gene set evaluation file.

4. The method of claim 3, wherein, The S2) comprises: integrating genes of the multiple predicted gene sets by using an EVM tool to obtain a preliminary integrated gene set; updating the preliminary integrated gene set by using Pasa software to obtain the integrated gene set.

5. The method of claim 4, wherein, The software for transcriptome prediction comprises Hisat2 software, Stringtie software, Trinity software, and Pasa software; the software for de novo prediction comprises one or more of August software, Snap software, or GeneMark; the software for homologous prediction is selected from genewise software.

6. An electronic device for genome annotation, comprising: The electronic device comprises a predicted gene set acquisition unit, an integrated gene set unit, a BUSCO evaluation unit, and a correction unit; wherein the predicted gene set acquisition unit is configured to predict gene structures of a genome in multiple ways to obtain multiple predicted gene sets; the integrated gene set unit is configured to integrate genes of the multiple predicted gene sets by using an EVM tool to obtain an integrated gene set; the BUSCO evaluation unit comprises a first BUSCO evaluation module and a second BUSCO evaluation module; the first BUSCO evaluation module is configured to perform BUSCO evaluation on the integrated gene set to obtain an integrated gene set evaluation file; the second BUSCO evaluation module is configured to perform BUSCO evaluation on a genome to obtain a genome evaluation file; The second BUSCO evaluation module is configured to perform BUSCO evaluation on the genome to obtain a genome evaluation file; The correction unit is configured to correct the integrated gene set evaluation file by using the genome evaluation file to obtain a corrected gene set. The gene structure prediction includes transcriptome prediction, de novo prediction and homology prediction. 7.The electronic device of claim 6, wherein, The correction unit includes a first correction module; The first correction module includes a first alignment unit and a first backfill unit. The first alignment unit is configured to align the genome evaluation file and the integrated gene set evaluation file to obtain a gene to be backfilled; The first backfill unit is configured to integrate the gene to be backfilled into the integrated gene set evaluation file; The gene to be backfilled is a gene that exists in the genome evaluation file but does not exist in the integrated gene set evaluation file. 8.The electronic device of claim 7, wherein, The correction unit further includes a second correction module; The second correction module includes a second alignment unit, a comparison unit, an integration unit, a removal unit and a retention unit. The second alignment unit is configured to align the genome evaluation file and the integrated gene set evaluation file to obtain the gene to be backfilled, as well as a first gene and a second gene; The comparison unit is configured to compare the sequence lengths of the first gene and the second gene; The integration unit is configured to integrate the first gene into the integrated gene set evaluation file when the sequence length of the first gene is longer than that of the second gene; The removal unit is configured to remove the second gene in the integrated gene set evaluation file when the sequence length of the first gene is longer than that of the second gene; The retention unit is configured to retain the second gene in the integrated gene set when the sequence length of the second gene is longer than that of the first gene; The first gene and the second gene have an intersection in the exon region; The first gene is derived from the genome evaluation file, and the second gene is derived from the integrated gene set evaluation file. 9.The electronic device of claim 8, wherein, The integrated gene set unit includes an integration module and a correction module; The integration module is configured to integrate genes by using an EVM tool on the plurality of predicted gene sets to obtain a preliminary integrated gene set; The correction module is configured to update the preliminary integrated gene set by using Pasa software to obtain the integrated gene set. 10.The electronic device of claim 9, wherein, The software for transcriptome prediction includes Hisat2 software, Stringtie software, Trinity software and Pasa software; The software for de novo prediction includes one or more of August software, Snap software or GeneMark; The software for homology prediction is selected from genewise software.

11. A computer readable storage medium, characterized in that, The storage medium includes a stored program, wherein when the program is running, the method for annotating a genome according to any one of claims 1-5 is executed.

12. A processor, comprising: The processor is used to run a program, wherein when the program is running, the method for annotating a genome according to any one of claims 1-5 is executed.

Images