Gene structure annotation method

Abstract

The invention provides a gene structure annotation method, and relates to the related field of biotechnology. A core gene set of a target species and a related species is obtained by comparing and identifying whole genome sequences of a researched species and a related species, and then the longest transcript sequence of a core gene is extracted and integrated into a transcriptome result so as to perfect and improve the quality of transcriptome data annotation of the target species. And a solid, reliable and comprehensive reference basis is provided for further improving the comprehensiveness and accuracy of gene structure annotation.

Description

Technical Field

[0001] The present invention relates to the technical field of bioinformatics analysis, and in particular to a method for annotating gene structure. Background Art

[0002] With the rapid development and gradual popularization of high-throughput sequencing technology, more and more organisms' genomes have been deciphered, from prokaryotic microorganisms to complex multicellular organisms. We have now entered a new era of "functional genomics", and genome annotation is the basis for studying the functions of various genes in organisms. Today, three methods, de novo prediction, homology prediction, and transcriptome data-assisted prediction, are widely used for gene structure annotation. Among them, transcriptome data is considered to be one of the important sources of gene structure annotation. However, due to the preciousness of some samples, the number of samples that can be collected is small, sequencing errors, insufficient sequencing depth, exogenous contamination, etc., the information of transcriptome data is often missing or its quality is not high enough, which directly affects the quality of gene structure annotation, resulting in the omission of some low-expressed genes and inaccurate gene positioning when revealing biological problems, which ultimately leads to unreliable biological conclusions.

[0003] At present, the method for gene structure annotation is mainly to integrate de novo prediction, homology prediction and transcriptome data through different weights. However, in the process of gene structure annotation, no method has been proposed to solve the problem of low quality of transcriptome data. Therefore, the proposed method of gene structure annotation has important significance and reference value.

[0004] In view of this, the present invention is proposed. Summary of the invention

[0005] In order to solve the above problems, the present invention provides a new method for gene structure annotation, which utilizes the principle of multiple sequence alignment, compares the whole genome sequences of closely related species and then identifies the core gene sequences, as well as screens a series of other conditions, to make up for the insufficient quality of the transcriptome data of the target species, fill in the source of evidence for gene annotation, and finally obtain high-quality gene structure annotation results.

[0006] In order to achieve the above-mentioned purpose, the present invention provides a method for gene structure annotation, which assembles the transcriptome data of the target species, then downloads the nucleic acid and protein amino acid sequences of closely related species from the public database for comparison and clustering, identifies the core sequences of closely related species, and then integrates the core gene series of closely related species of the same genus and species into the assembled transcriptome data, thereby obtaining a high-quality reference data set as a source of evidence for gene structure annotation, and finally obtaining a comprehensive and accurate gene structure annotation result. The method of the present invention is carried out by the following steps:

[0007] Step 1. Transcriptome assembly: Assemble the transcriptome data of the target species using Trinity software;

[0008] Step 2. Data preparation: Download the nucleic acid sequence of the longest transcript of the target species' closely related species and the amino acid sequence file of the encoded protein from public databases such as NCBI and CNGB;

[0009] Step 3. Gene family clustering: Use OrthoFinder software to perform gene family clustering analysis on the gene sets of homologous species, and identify the orthologous gene pairs of closely related species through the gene family clustering results;

[0010] Step 4. Determine whether the closely related species in step 2 are closely related species of the same genus and species as the target species;

[0011] Step 5. Define the common orthologous gene pairs in the genomes of closely related species screened in step 3 as the core gene set, and select the longest transcript of the closest related species as the candidate reference sequence;

[0012] Step 6. If there are closely related species of the same genus and species, the longest transcript sequence in the closely related species of the same genus and species is integrated into the transcriptome assembly result of step 1; if there are no closely related species of the same genus and species, the ancestral consensus sequence of all species is predicted using PAML software, and the ancestral sequence is extracted as the reference sequence, which is then merged with the transcriptome result of step 1 to obtain the reference transcript sequence for auxiliary annotation of the target species;

[0013] Step 7. Use PASA (Program to Assemble Spliced ​​Alignments) software to annotate the transcript set obtained in step 6 to make up for the incompleteness or low quality of transcriptome data;

[0014] Step 8. Using the gene structure obtained in step 7 as the skeleton, use EVM (EVidence Modeler) software, set reasonable weight values, integrate the results of de novo prediction, homology annotation, and transcriptome annotation, and obtain high-quality gene structure annotation results.

[0015] In the example of the present invention, the environment configuration required for the method of using software to annotate gene structure is: the hardware configuration is a PC client with a Pentium II or higher CPU, 16G or higher memory, and 80G or higher hard disk; the software configuration is a Linux operating system and is configured with gcc 4.9.3 and above, Python 2.7.9 and above, Perl 5.22.0 and above, R2.15.3 and above, Orthofinder (or OrthMCL, TreeFam) any version, PAML software package, de novo prediction software August, GeneScan, Glimmer, etc., homology annotation software Genewise, transcriptome auxiliary annotation software PASA, integrated annotation software EVM.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] The method for gene structure annotation provided by the present invention is based on a new gene structure annotation strategy, which can effectively make up for the negative impact on genome annotation caused by incomplete transcriptome data information due to various reasons, and can more comprehensively identify all genes on the genome, improve the accuracy and completeness of gene structure annotation, and provide comprehensive and important reference value and technical support for further comprehensively revealing various genes and their functions in biological organisms. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the main contents and specific implementation methods of the present invention, Figure 1 The specific implementation modes of the present invention are briefly introduced, wherein Trinity, Cufflinks, OrthoFinder, PAML, EVM, and Pasa are bioinformatics analysis software; pep is the abbreviation of amino acid sequence; and CDS is the abbreviation of nucleic acid sequence.

[0019] Figure 1 The technical flow chart provided for the present invention. DETAILED DESCRIPTION

[0020] The embodiments of the present invention will be described in detail below in combination with implementation methods and specific implementation cases, but those skilled in the art will understand that the following implementation methods and examples are only used to illustrate the present invention and should not be construed as limiting the scope of the present invention.

[0021] The present invention provides a method for gene structure annotation, which is performed by the following steps:

[0022] Step 1. Transcriptome assembly: Assemble the transcriptome data of the target species using Trinity software;

[0023] Step 2. Data preparation: Download the nucleic acid sequence of the longest transcript of the target species' closely related species and the amino acid sequence file of the encoded protein from public databases such as NCBI and CNGB;

[0024] Step 3. Gene family clustering: Use OrthoFinder software to perform gene family clustering analysis on the gene sets of homologous species, and identify the orthologous gene pairs of closely related species through the gene family clustering results;

[0025] Step 4. Determine whether the closely related species in step 2 are closely related species of the same genus and species as the target species;

[0026] Step 5. Define the common orthologous gene pairs in the genomes of closely related species screened in step 3 as the core gene set, and select the longest transcript of the closest related species as the candidate reference sequence;

[0027] Step 6. If there are closely related species of the same genus and species, the longest transcript sequence in the closely related species of the same genus and species is integrated into the transcriptome assembly result of step 1; if there are no closely related species of the same genus and species, the ancestral consensus sequence of all species is predicted using PAML software, and the ancestral sequence is extracted as the reference sequence, which is then merged with the transcriptome result of step 1 to obtain the reference transcript sequence for auxiliary annotation of the target species;

[0028] Step 7. Use PASA (Program to Assemble Spliced ​​Alignments) to annotate the transcript set obtained in step 6 to make up for the incompleteness or low quality of transcriptome data;

[0029] Step 8. Using the gene structure obtained in step 7 as the skeleton, use EVM (EVidence Modeler) software, set reasonable weight values, integrate the results of de novo prediction, homology annotation, and transcriptome annotation, and obtain high-quality gene structure annotation results.

[0030] In some optional implementation steps, the software for performing gene family clustering in step 3 may also use OrthoMCL or Treefam.

[0031] In some optional implementation steps, the prediction of the ancestral sequence in step 6 adopts the ancester model.

[0032] In some optional implementation steps, the software for integrating the de novo prediction, homology annotation and transcriptome annotation data in step 8 may also use BRAKER or MAKER3.

[0033] In some optional implementation steps, the homology annotation software for plants in step 8 is Exonerate.

[0034] In some optional implementation steps, step 8 also includes: using PASA software to upgrade the annotation results obtained in step 8 to obtain higher quality gene structure annotation results.

[0035] The present invention is based on the principle of whole genome comparison of closely related species. The nucleic acid sequences of the longest transcripts and the amino acid sequences encoding proteins of multiple closely related species with a target species are selected for comparison to identify the core gene sequences of closely related species, which are used as a source of reliable evidence for auxiliary annotation of gene structure and make up for the lack of transcriptome data of the target species. Compared with the existing gene structure annotation method which only integrates annotation information from three different sources, namely, de novo prediction, homology prediction and transcriptome data, the gene sequences of homologous species are used to supplement and improve the existing transcriptome data of the target species, which greatly improves the accuracy, completeness and comprehensiveness of gene structure annotation.

[0036] In the embodiments of the present invention, the quality of gene structure annotation can be further improved by compensating for the defects of the transcriptome data of the target species. The present invention has been verified in practice in the gene structure annotation of a ruminant and a rodent, and the present invention is further illustrated by specific embodiments below.

[0037] Example 1

[0038] Taking the gene structure annotation of a ruminant as an example, the annotation process is as follows:

[0039] Ruminants are identified as mammals, and their closely related species are identified as mammal-A, mammal-B, mammal-C, and mammal-D.

[0040] 1) Use Trinity software to assemble mammal transcriptome data;

[0041] 2) Download the amino acid sequence mammal-A-pep and nucleic acid sequence mammal-A-CDS of the female genome of the ruminant from NCBI, as well as the amino acid sequence and nucleic acid sequence of its three closely related species, namely mammal-B-pep, mammal-C-pep, mammal-D-pep, mammal-B-CDS, mammal-C-CDS, and mammal-D-CDS;

[0042] 3) using the software OrothoFinder to cluster the homologous gene sets of the species described in 2) and identify the orthologous gene pairs shared by closely related species;

[0043] 4) selecting the longest transcript of the core gene of the female ruminant from 3) as a candidate reference sequence;

[0044] 5) integrating the transcriptome assembly results obtained in 1);

[0045] 6) Use PASA to annotate gene structure based on transcript evidence obtained in 5);

[0046] 7) Using the gene structure predicted in 6) as the basic framework, the EVM software was used to integrate the de novo prediction, homology prediction and the improved transcriptome annotation results obtained above for the species to obtain the annotation results of the gene structure.

[0047] The quality of gene structure annotation was evaluated according to the above analysis method. Using the traditional method, the results from de novo prediction, homology prediction and transcriptome annotation were integrated using EVM software, and a total of 25,932 protein-coding genes were annotated, with a BUSCO evaluation of 79.7% (C: 79.7% [S: 77.5%, D: 2.2%], F: 6.7%, M: 13.6%, n: 3354). The transcriptome data of the ruminant was further improved by the present invention, and the results of gene structure annotation were improved. A total of 26,294 protein-coding genes were identified, and the BUSCO evaluation was 92.1% (C: 92.1% [S: 89.7%, D: 2.4%], F: 3.0%, M: 4.9%, n: 3354). The above examples show that the results of the present invention are reliable and can effectively improve the accuracy and completeness of ruminant gene structure annotation.

[0048] Example 2

[0049] Taking the gene structure annotation of a rodent as an example, the annotation process is as follows:

[0050] Rodents are identified as Rodent, and their closely related species are identified as rhizomys_sinensis, bamboo_rat, and Rhizomys_pruinosus.

[0051] 1) Assemble Rodent transcriptome data using Trinity software;

[0052] 2) Download the amino acid sequences of rhizomys_sinensis, bamboo_rat and Rhizomys_pruinosus from NCBI: rhizomys_sinensis_pep, bamboo_rat_pep, Rhizomys_pruinosus_pep and the nucleic acid sequences of rhizomys_sinensis_cds, bamboo_rat_cds, Rhizomys_pruinosus_cds;

[0053] 3) using the software OrthoFinder to cluster the homologous gene sets of the species described in 2) and identify the orthologous gene pairs of closely related species;

[0054] 4) selecting the longest transcript of rhizomys_sinensis, a closely related species of the target rodent, from 3) as a candidate reference sequence;

[0055] 5) integrating the transcriptome assembly results obtained in 1);

[0056] 6) using PASA to annotate the combination of transcripts obtained in 5);

[0057] 7) Using the gene structure of 6) as the skeleton, the EVM software was used to integrate the de novo prediction, homology prediction and the improved transcriptome annotation results obtained above for the species to obtain the annotation results of the gene structure of the rodent.

[0058] The quality of gene structure annotation was evaluated according to the above analysis method. Using the traditional method, the results from de novo prediction, homology prediction and transcriptome annotation were integrated using EVM software, and a total of 31,465 protein-coding genes were annotated, with a BUSCO evaluation of 64.3% (C: 64.3% [S: 58.9%, D: 5.4%], F: 14.3%, M: 21.4%, n: 3354). The transcriptome data of the ruminant was further improved by the present invention, and the results of gene structure annotation were improved. A total of 39,742 protein-coding genes were identified, and the BUSCO evaluation was 89.0% (C: 89.0% [S: 80.8%, D: 8.2%], F: 5.3%, M: 5.7%, n: 3354). The above studies show that the present invention greatly improves the integrity of gene structure annotation, and has universal applicability, and is suitable for gene structure annotation research of various species.

[0059] Note: The gene numbers in the above two examples are the evaluation results after integration, and the genes in the repeated sequence region have not been filtered, that is, the results after PASA update.

[0060] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some or all of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for annotating gene structure, characterized in that: Follow these steps: Step 1. Transcriptome assembly: Assemble the transcriptome data of the target species using Trinity software; Step 2. Data preparation: Download the nucleic acid sequence of the longest transcript of the target species' closely related species and the amino acid sequence file of the encoded protein from public databases such as NCBI and CNGB; Step 3. Gene family clustering: Use OrthoFinder software to perform gene family clustering analysis on the gene sets of homologous species, and identify the orthologous gene pairs of closely related species through the gene family clustering results; Step 4. Determine whether the closely related species in step 2 are closely related species of the same genus and species as the target species; Step 5. Define the common orthologous gene pairs in the genomes of closely related species screened in step 3 as the core gene set, and select the longest transcript of the closest related species as the candidate reference sequence; Step 6. If there are closely related species of the same genus and species, the longest transcript sequence in the closely related species of the same genus and species is integrated into the transcriptome assembly result of step 1; If there are no closely related species of the same genus and species, use PAML software to predict the ancestral consensus sequence of all species, extract the ancestral sequence as the reference sequence, and then merge it with the transcriptome results in step 1 to obtain the reference transcript sequence for auxiliary annotation of the target species; Step 7. Use PASA (Program to Assemble Spliced ​​Alignments) software to annotate the transcript set obtained in step 6 to make up for the incompleteness or low quality of transcriptome data; Step 8. Using the gene structure obtained in step 7 as the skeleton, use EVM (EVidence Modeler) software, set reasonable weight values, integrate the de novo prediction, homology annotation, and transcriptome annotation results to obtain high-quality gene structure annotation results.

2. The method according to claim 1, characterized in that In step 3, software for gene family clustering can also be selected such as OrthMCL or Treefam.

3. The method according to claim 1, characterized in that In step 6, the ancestral sequence is predicted using the ancester model.

4. The method according to claim 1, characterized in that The software used to integrate the de novo prediction, homology annotation and transcriptome annotation data in step 8 is BRAKER or MAKER3.

5. The method according to claim 1, characterized in that The environment configuration required for the method of using software to annotate gene structure is: the hardware configuration is a PC client with a Pentium II or higher CPU, 16G or higher memory, and 80G or higher hard disk; the software configuration is a Linux operating system and is configured with gcc 4.9.3 and above, Python 2.7.9 and above, Perl5.22.0 and above, R 2.15.3 and above, Orthofinder (or OrthoMCL, TreeFam) any version, PAML software package, de novo prediction software August, GeneScan, Glimmer, etc., homology annotation software Genewise, transcriptome auxiliary annotation software PASA, integrated annotation software EVM.

Images