Zhiwen Yang

Zhiwen Yang

I have a wide range of research interests, including machine learning, deep learning and their application to computational biology, with a focus on developing tools to solve real world problems.

PRISM: A Proteomics Robust Imputation framework for Structure-aware Modeling of missingness

Published in Nature Communications (under review, preprint), 2025

This work develops PRISM, a robust imputation framework for proteomics data with missing-not-at-random (MNAR) values.

Key contributions include:

  • Integrates a Denoising Convolutional Autoencoder (DCAE) with Deep Matrix Factorization (DMF) to jointly model protein intensity patterns and missingness structure.
  • Achieves improved imputation accuracy over classical methods (e.g., KNN, MissForest, Gaussian-based approaches) while preserving biological signal.
  • Demonstrates that PRISM better maintains downstream biological structure, supporting more reliable differential expression and pathway analysis.

Recommended citation: Li, Z., Yang, Z., Chen, Y., & Guo, T. (2025). PRISM: A Proteomics Robust Imputation framework for Structure-aware Modeling of missingness. Nature Communications (under review, preprint).
Download Paper