Yang Li

Estimand-aware and donor-aware triangulation of predefined gene-set signals across human tendon transcriptomic datasets

The paper analyzes gene-set signals across three human tendon transcriptomic datasets to identify consistent biological differences and highlights the importance of integrating data by estimand and biological unit.

scKEPLM: Knowledge enhanced large-scale pre-trained language model for single-cell transcriptomics

scKEPLM integrates a biology knowledge graph into single-cell transcriptome pre-training to improve gene and cell representation, achieving state-of-the-art performance in multiple downstream tasks.