Xingyu Liao

FastSTR efficiently identifies and characterizes tandem repeats across 13 species with high accuracy and speed, uncovering new genomic elements and disease-related variations.

This paper introduces a deep learning method to classify CRISPR-Cas systems using repeat sequences, improving classification accuracy for subtypes with limited data.

BERTE improves TE classification accuracy using BERT and CNN, achieving up to 21% higher F1-score than existing methods.

This paper introduces CREATE, a deep learning framework for TE classification, and WebDLTE, a web platform for real-time TE annotation, showing improved performance over existing tools.

The paper introduces AFS, a tool to revoke patients' private data from pre-trained models by auditing usage and purifying knowledge, tested across multiple tasks and datasets.

DeepTRs identifies tandem repeat variations directly from Nanopore reads using multi-modal conversion and deep learning, achieving high sensitivity, accuracy, and completeness.

AB-Gen uses reinforcement learning and a generative pre-trained Transformer to design antibody sequences with multiple desirable properties, improving success rates over traditional methods.

The paper accelerates RepeatClassifier using Spark and a greedy algorithm with dynamic upper boundary, achieving 42-88x speedup while maintaining accuracy.