HFTC: A Hierarchical Fungal Taxonomic Classification Model for ITS Sequences Using Low-Dimensional Embedding Features

Jiawei Wang¹Shaojie Qiao^1*Chao Wang^2*Dongsheng Xiang¹Yangcheng Liao¹

• ¹Chengdu University of Information Technology, Chengdu, China
• ²Guangxi Medical University, Nanning, China

Introduction: Fungal identification through ITS sequencing is pivotal for biodiversity and ecological studies, yet existing methods often face challenges with high-dimensional features and inconsistent taxonomy predictions. Method: We proposed HFTC, a hierarchical fungal taxonomic classifier built upon a multi-level Random Forest architecture. Notably, HFTC incorporates a bidirectional k-mer strategy to capture contextual information from both sequence orientations. By leveraging Word2Vec embedding, it reduces the feature dimensionality from 4k to just 200, significantly improving computational efficiency while preserving rich sequence context. Result: Experimental results demonstrate that HFTC outperforms Mothur, RDP, Sintax, QIIME2 and CNN-Duong, achieving MCC of 95.31% despite uneven class distributions. Its overall ACC reaches 95.25%. Notably, it attains a hierarchical accuracy (HA) of 95.10% at the species level, surpassing the best-performing deep learning baseline, CNN-Duong, by 3.2%. Moreover, HFTC exhibits the smallest discrepancy between ACC and HA for just 1.60‰, in contrast to CNN-Duong, which shows the largest gap for 35.00‰, highlighting HFTC's superior hierarchical consistency. Discussion: HFTC offers a scalable and accurate approach for fungal taxonomic classification. Its compact feature representation and hierarchical architecture make it particularly suitable for microbial diversity research.