物种准确鉴定是生物学研究的基石。目前，真菌鉴定主要依赖形态学鉴定和DNA条形码鉴定，但这些方法在近缘物种鉴定方面存在一定局限性。本论文以青霉属7个具有重要经济价值的近缘物种为例，建立基于全基因组分析（Analysis of whole-GEnome，AGE）的近缘物种鉴定体系；进一步在青霉属82个近缘物种中验证AGE的适用性；最后应用AGE解决镰刀菌属近缘物种鉴定难题。主要研究结果如下：

1. 建立青霉属7个近缘物种鉴定的AGE方法。本研究为青霉属真菌近缘物种准确鉴定提出AGE方法。AGE的操作流程包括：通过生物信息学分析从目标物种全基因组中筛选特异靶标序列，利用实验技术检测特异靶标序列，从而实现物种鉴定。以青霉属7个近缘物种为例：变灰青霉（Penicillium canescens）、草酸青霉（P. oxalicum）、桔青霉（P. citrinum）、酒青霉（P. paneum）、娄地青霉（P. roqueforti）、产红青霉（P. rubens）和波兰青霉（P. polonicum），通过生物信息学分析其基因组数据，构建7个物种特异靶标序列库，去重后特异靶标序列数量范围为24,948至772,043，从特异靶标序列库中为7个物种各筛选一条特异靶标序列，利用Sanger测序和CRISPR-Cas12a技术成功实现7个物种的准确鉴定。AGE灵敏度低至0.01 ng/µL，且可通过荧光值、可视荧光和试纸条三种形式呈现鉴定结果，显著降低对仪器的依赖，具备现场检测能力。本研究以青霉属7个近缘物种为研究对象，建立真菌近缘物种鉴定的AGE方法，为后续研究奠定基础。

2. 利用AGE构建青霉属82个近缘物种特异靶标序列库，实现物种准确鉴定。物种分析范围不同，AGE筛选的物种特异靶标序列可能发生变化，在开展相关研究时需纳入更多的物种基因组数据进行分析。研究发现根据青霉属7个近缘物种筛选的实验验证特异靶标序列在青霉属82个近缘物种中不具备特异性，为提高特异靶标序列的适用范围，本研究覆盖青霉属82个物种共130个基因组数据。通过生物信息学分析构建82个物种特异靶标序列库，去重后特异靶标序列数量范围为2,562至557,355条。从特异靶标序列库中为青霉属7个近缘物种（P. canescens、P. oxalicum、P. citrinum、P. paneum、P. roqueforti、P. rubens和P. polonicum）各筛选一条特异靶标序列。为验证筛选的物种特异靶标序列的可靠性，利用Sanger测序和CRISPR-Cas12a技术进行检测，结果表明这7条特异靶标序列具有高特异性和适用性，尤其是准确检测了ITS不易鉴定的P. canescens、P. citrinum和P. rubens。值得注意的是，来自基因组未注释区域的特异靶标序列在物种鉴定中展现出很好的分辨能力。此外，为解决P. citrinum的近缘物种P. hetheringtonii以及P. rubens的近缘物种P. chrysogenum鉴定难题，通过生物信息学分析分别为其筛选一条特异靶标序列，为它们的准确鉴定提供参考。以上结果表明，通过对青霉属82个近缘物种的生物信息学分析，AGE能够实现该属82个近缘物种的准确鉴定，为解决其他真菌属的近缘物种鉴定难题提供借鉴。

3. 利用AGE构建镰刀菌属173个近缘物种特异靶标序列库，为物种鉴定难题提供解决方案。镰刀菌属真菌一直是真菌物种鉴定的重点和难点，传统形态鉴定和现有的分子鉴定方法都存在不易准确区分其近缘物种的问题。使用生物信息学对镰刀菌属173个物种共304个基因组数据进行分析，成功构建173个物种的特异靶标序列库，特异靶标序列数量范围为102到409,550条之间。从特异靶标序列库中为具有重要临床和农业意义的镰刀菌属7个物种（尖孢镰刀菌（Fusarium oxysporum）、腐皮镰孢菌（F. solani）、藤仓镰刀菌（F. fujikuroi）、接骨木镰孢菌（F. sambucinum）、黄色镰孢菌（F. culmorum）、肉红镰孢菌（F. incarnatum）和木贼镰刀菌（F. equiseti））各选取一条特异靶标序列。利用Sanger测序和CRISPR-Cas12a技术对7个物种的特异靶标序列进行验证，结果表明筛选的7个物种特异靶标序列具有高特异性，不仅准确鉴定了常见的致病性物种，还解决了条形码难以准确区分的近缘物种F. incarnatum和F. equiseti鉴定难题。此外，为解决F. graminearum种属群高度近缘物种（F. graminearum、F. asiaticum和F. meridionale）的鉴定难题，分别从特异靶标序列库中为其各挑选一条特异靶标序列，为实现物种准确鉴定提供参考。总之，AGE能够实现镰刀菌属近缘物种在物种水平的准确鉴定，为该属的多种密切相关复合物种的准确鉴定提供解决方案。

综上所述，本论文基于青霉属7个物种建立真菌近缘物种鉴定的AGE方法，并通过实验技术证明方法的可行性；扩大青霉属近缘物种分析范围后，筛选的物种特异靶标序列在近缘物种鉴定中展现出很好的适用性；AGE可解决镰刀菌属近缘物种鉴定难题，为真菌物种鉴定提供重要工具。

Species identification is a cornerstone of biological research. Currently, fungal identification primarily relies on morphological identification and DNA barcoding, but these methods have certain limitations in distinguishing closely related species. This study establishes a species identification system based on Analysis of whole-GEnome (AGE) using seven economically important closely related species of the genus Penicillium as examples. The applicability of AGE was further verified in 82 closely related species of Penicillium. Finally, AGE was applied to solve the challenges of identifying closely related species in the genus Fusarium. The main research results are as follows:

1. An AGE method is established for the identification of closely related fungal species based on seven Penicillium species. This study proposes an AGE method for the accurate identification of closely related species in Penicillium. The operational workflow of AGE includes: selecting species-specific target sequences from whole-genome data of target species through bioinformatics analysis, detecting these sequences using experimental techniques, and thereby achieving species identification. Using seven closely related species of Penicillium (P. canescens, P. oxalicum, P. citrinum, P. paneum, P. roqueforti, P. rubens, and P. polonicum) as examples, bioinformatics analysis of their genomic data was conducted to construct  species-specific target sequence libraries for the seven species. After deduplication, the number of species-specific target sequences ranged from 24,948 to 772,043. One species-specific sequence was selected from the library for each of the seven species, and Sanger sequencing and CRISPR-Cas12a technology were successfully used to achieve accurate identification of the seven species. The sensitivity of AGE was as low as 0.01 ng/µL, and results could be presented in three forms: fluorescence values, visible fluorescence, and lateral flow strip. This significantly reduced dependence on instrumentation and enabled on-site detection. This study used seven species of Penicillium as research objects to establish an AGE method for the identification of closely related fungal species, laying the foundation for future research.

2. Construction of species-specific target sequence libraries for 82 closely related species of the genus Penicillium by AGE enables accurate species identification. The number of species-specific sequences selected using AGE may vary depending on the scope of species analysis. Therefore, more genomic data from additional species should be included in the analysis for related studies. It was found that the experimentally validated species-specific target sequences selected from the seven closely related species of Penicillium did not exhibit specificity when tested across 82 species of Penicillium. To expand the applicability of species-specific target sequences, this study included genomic data from 130 genomes covering 82 species of Penicillium. Species-specific target sequence libraries were constructed through bioinformatics analysis, with sequences ranging from 2,562 to 557,355 after deduplication. One species-specific target sequence was selected from the libraries for each of the seven closely related species of Penicillium (P. canescens, P. oxalicum, P. citrinum, P. paneum, P. roqueforti, P. rubens, and P. polonicum). To verify the reliability of the selected species-specific target sequences, Sanger sequencing and CRISPR-Cas12a technology were used for detection. The results showed that these seven target sequences exhibited high specificity and applicability, particularly in accurately identifying P. canescens, which is difficult to distinguish using ITS barcoding. Notably, species-specific target sequences derived from unannotated regions of the genome demonstrated excellent discriminatory power in species identification. Furthermore, to address the challenges of identifying closely related species such as P. hetheringtonii (closely related to P. citrinum) and P. chrysogenum (closely related to P. rubens), bioinformatics analysis was used to select one species-specific target sequence for each, providing a reference for their accurate identification. These results indicate that AGE can achieve accurate identification of 82 closely related species of Penicillium through bioinformatics analysis, providing important insights for resolving challenges in identifying closely related species in other fungal genera.

3. Construction of specific target sequence libraries for 173 closely related species of the genus Fusarium by AGE provides solution to species identification issue. Species of the genus Fusarium have long been a focus and challenge in fungal species identification, with traditional morphological and existing molecular identification methods struggling to accurately distinguish closely related species. Bioinformatics analysis of 304 genomes from 173 species of Fusarium was conducted, successfully constructing species-specific target sequence libraries for the 173 species, with sequence numbers ranging from 102 to 409,550. One species-specific target sequence was selected from the libraries for each of seven species of Fusarium with significant clinical and agricultural importance (F. oxysporum, F. solani, F. fujikuroi, F. sambucinum, F. culmorum, F. incarnatum, and F. equiseti). Sanger sequencing and CRISPR-Cas12a technology were used to validate the selected species-specific target sequences, demonstrating that the seven sequences exhibited high specificity. They not only accurately identified common pathogenic species but also resolved the challenge of distinguishing the closely related species F. incarnatum and F. equiseti, which are difficult to differentiate using barcoding. Additionally, to address the challenge of distinguishing the highly related species within the F. graminearum species complex (F. graminearum, F. asiaticum, and F. meridionale), a specific target sequence was selected from the specific target sequence libraries for each species, providing a reference for their accurate identification. In summary, AGE can address the challenges of identifying closely related species in the genus Fusarium, offering solutions for the accurate identification of multiple closely related species within the genus.

In conclusion, this study established an AGE method for the identification of closely related species using seven species of Penicillium as examples and demonstrated the feasibility of the method through experimental validation. Expanding the scope of Penicillium species analysis showed that the selected species-specific target sequences exhibited excellent applicability in identifying closely related species. Furthermore, AGE successfully addressed the challenges of identifying closely related species in the genus Fusarium, providing an important tool for fungal species identification.