Marc XU

• Independently developed a consensus scoring algorithm that improved structure-based virtual screening accuracy by 10–20%. • Led cross-border collaboration between SIAT (China) and ETH Zurich (Switzerland) to validate the pairwise consensus scoring method through in vitro experiments. • Authored and co-authored multiple publications contributing to the development of novel drug candidates targeting GPCRs. • Supported preparation of supervisor’s research presentations for international conferences.

• Played a key role in end-to-end drug discovery projects, covering target identification, virtual screening, ADMET prediction, and clinical candidate selection. • Developed AI models for predicting 15 key ADMET properties with over 80% accuracy. • Designed and implemented a GPCR–signal pathway–disease knowledge graph to support and accelerate bioinformatics analysis. • Bridged R&D and application by integrating in silico drug discovery platforms into practical workflows. • Standardized database template and trained internal teams for consistent data management and usage. • Achieved top 3 ranking in the GPCR Dock 2021 international competition; model accuracy surpassed Google AlphaFold by over 60%. • Played a key role in full-stack IT development, spanning frontend to backend, and authored patent applications. • Developed a computational platform for in silico drug discovery and optimization druglike properties.

• Developed a PSSM-based tool for accurate and automated annotation of GPCR families and subfamilies directly from raw sequences, from transmembrane domain detection to mapping. • Coordinated a collaborative research project between international teams from China and Finland.

• Conducted in-depth analyses of NS1 protein sequences to identify structurally and functionally targetable regions for novel influenza therapies. • Translated complex biological requirements into precise and efficient computational solutions.