Ученые создали автоматизированную систему для анализа морфологии промышленных грибов на всех этапах развития

This study was led by Dr. Xiaomei Zheng, Prof. Ping Zheng, and Prof. Sun (Tianjin Institute of Industrial Biotechnology, Chinese Academy of Sciences). Industrial filamentous fungi, such as Aspergillus niger, are widely used for the production of citric acid, enzymes, and recombinant proteins. Their fermentation performance is strongly influenced by fungal morphology throughout development, from spore germination and hyphal growth to mature pellet formation. However, existing morphology-analysis methods are largely restricted to specific developmental stages and provide only limited quantitative information, hindering systematic investigation of morphology–productivity relationships. To address this challenge, the authors developed PAMP (Python-based Automated Morphological Analysis Pipeline), an open-source image-analysis platform capable of automatically quantifying fungal morphology throughout the entire life cycle. By integrating adaptive image segmentation, skeleton extraction, and contour reconstruction, PAMP enables high-resolution characterization of spore germination, hyphal networks, and pellet architecture within a unified analytical framework. The platform also supports fully automated batch processing, allowing rapid analysis of thousands of microscopic images with standardized quantitative outputs. The authors first validated the quantitative accuracy of PAMP by comparing automated measurements with independent manual measurements performed in ImageJ. Across representative morphological parameters, including germ tube length, hyphal length, branching frequency, pellet core radius, peripheral hyphal length, and pellet radius, PAMP showed excellent agreement with manual measurements, demonstrating its robustness and reliability for automated fungal morphology analysis. The team further applied PAMP to quantitatively characterize developmental differences between industrial A. niger strains and a conditional pkaC expression mutant. The analyses revealed that early hyphal growth patterns strongly influence subsequent pellet architecture. Compared with conventional morphology-analysis workflows, PAMP more accurately preserved low-contrast peripheral hyphae and enabled quantitative characterization of pellet core, peripheral hyphae, and pellet porosity, providing structural descriptors that were previously inaccessible. Importantly, correlation analyses demonstrated that peripheral architectural features quantified by PAMP, particularly peripheral hyphal length and pellet porosity, exhibited substantially stronger associations with citric acid production than conventional pellet descriptors. These findings reveal that the peripheral region of fungal pellets represents the major functional zone governing mass transfer and fermentation performance, providing new insights into the relationship between fungal morphogenesis and industrial productivity. "Our goal was not only to automate fungal morphology analysis, but also to establish a quantitative framework linking fungal development with fermentation performance", says Dr. Zheng. "We hope that PAMP will serve as an open-source platform for morphology-guided strain engineering and facilitate the application of artificial intelligence in industrial biotechnology". The study establishes a standardized, high-throughput and high-resolution framework for fungal morphological phenotyping. By enabling quantitative analysis across the complete fungal life cycle, PAMP provides a valuable resource for investigating morphogenesis, optimizing industrial fermentation, and accelerating intelligent strain design in filamentous fungi. See the article: PAMP: an automated high-resolution image analysis pipeline for life-cycle quantitative morphological profiling in filamentous fungi https://doi.org/10.1080/21501203.2026.2710468 Founders - The work was supported by the Key R&D Program of Shandong Province (2022SFC0101), Strategic Priority Research Program of the Chinese Academy of Sciences (XDA0510300).