SC-AIR-BERT: a pre-trained single-cell model for predicting the antigen-binding specificity of the adaptive immune receptor.

Accurately predicting the antigen-binding specificity of adaptive immune receptors (AIRs), such as T-cell receptors (TCRs) and B-cell receptors (BCRs), is essential for discovering new immune therapies. However, the diversity of AIR chain sequences limits the accuracy of current prediction methods. This study introduces SC-AIR-BERT, a pre-trained model that learns comprehensive sequence representations of paired AIR chains to improve binding specificity prediction. SC-AIR-BERT first learns the 'language' of AIR sequences through self-supervised pre-training on a large cohort of paired AIR chains from multiple single-cell resources. The model is then fine-tuned with a multilayer perceptron head for binding specificity prediction, employing the K-mer strategy to enhance sequence representation learning. Extensive experiments demonstrate the superior AUC performance of SC-AIR-BERT compared with current methods for TCR- and BCR-binding specificity prediction.

[Characteristics of Eukaryotic Phytoplankton Community Structure and Its Relationship with Environmental Factors in Danjiangkou Reservoir].

Eukaryotic phytoplankton play crucial roles in ecosystem processes, and understanding their community composition and abundance is of great importance to protect the balance of aquatic ecosystems. Danjiangkou Reservoir is a canyon lake in central China that acts as the water source of the Mid-route of South-to-North Water Diversion Project. In this study, the composition of planktonic microeukaryotes from water samples with a depth ranging from 0.5 to 20 m in Danjiangkou Reservoir were investigated using high-throughput 18S rDNA gene sequencing. The environmental factors including pH, water temperature (WT), nitrate (NO3--N), ammonia (NH4+-N), total nitrogen (TN), conductivity (Cond.), water depth, and dissolved oxygen (DO) in the reservoir areas were measured, and their correlations with the abundance and diversity of eukaryotic phytoplankton were analyzed. The results of high-throughput sequencing showed that a total of 68 genera of phytoplankton belonging to 12 phyla were detected, and the phytoplankton communities were mainly composed of Dinophyceae, Bacillariophyta, and Chlorophyta. In particular, significant differences in the diversity of phytoplankton communities on a vertical distribution were found. The diversity and abundance of the phytoplankton communities in Songgang (SG), Tumen (TM), and Dangzikou (DZK) decreased with the increase in water depth, and the relationship between phytoplankton and water depth in other sampling sites was not obvious. The results of RDA showed that NO3--N, water depth, DO, pH, and WT could influence the vertical distribution of the phytoplankton community in the Danjiangkou Reservoir. Among these phytoplankton types, Dinophyceae was negatively correlated with NO3--N and water depth and positively correlated with other environmental factors. Meanwhile, Bacillariophyta was positively correlated with NO3--N and water depth and negatively correlated with other environmental factors. In addition, Chlorophyta was negatively correlated with WT, pH, and DO and not significantly correlated with water depth.