Biologically informative not available deconvolution excavates hidden features in proteomics datasets
Briefings in Bioinformatics · 2026
DOI: 10.1093/bib/bbag582 ↗Kurz gefasst
Fachartikel in der Zeitschrift Briefings in Bioinformatics, 2026. Verfasst von Weiheng Guo und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biomedical Sciences, College of Biomedicine, City University of Hong Kong, Tat Chee Ave, Kowloon, Hong Kong SAR 999077, China. Hauptthemen laut MEDLINE: Proteomics, Databases, Protein, Liver Neoplasms, Carcinoma, Hepatocellular. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Institute of Digital Medicine, Tung Foundation Biomedical Sciences Centre, Research Grants Council of Hong Kong.
Abstract (Originalsprache)
Missing values [not available (NAs)] in proteomics arise from true absence/low abundance and technical factors. Conventional imputation often obscures disease signals by ignoring the biological information inherent in protein absence. We present biologically informative NA deconvolution (BIND), an adaptive neighborhood-based workflow that classifies NAs as biological or technical. BIND binarizes matrices to encode presence/absence and uses context-aware k-nearest neighbors with Gaussian kernels to deconvolve NA origins, enabling clustering, biomarker discovery, and an NA-weighted correlation for protein–protein interaction (PPI) inference. In a simulation derived from the CCLE949 dataset, BIND achieved an Area Under the Curve (AUC) of 0.93, with k = 10 and γ = 3 selected for subsequent Cancer Cell Line Encyclopedia (CCLE) analyses, and remained robust across biological-to-technical NA ratios ranging from 3:1 to 1:3. In hepatocellular carcinoma tissues (n = 248), BIND distinguished tumor from adjacent normal, identified candidate biomarkers, and detected cancer-specific PPIs enriched in spliceosome and DNA replication pathways. A public BIND web server offers interactive visualization and supports single-cell proteomics, enhancing accessibility. Overall, BIND transforms NA patterns from a liability into biologically informative features for clustering, endotype discovery, biomarker identification, and PPI analysis.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Biomedical Sciences, College of Biomedicine, City University of Hong Kong, Tat Chee Ave, Kowloon, Hong Kong SAR 999077, China.
- Department of Orthopaedics, Renmin Hospital of Wuhan University, Wuhan University, Zhangzhidong Road, Wuhan 430060, China.
- School of Software, Northwestern Polytechnical University, West Youyi Road, Xi'an 710072, China.
- Department of Pharmaceutical and Artificial-Intelligence Sciences, School of Medicine, Shanghai Jiao Tong University, South Chongqing Road, Shanghai 200025, China.
ORCID-Profile: Jian Zhang, Xikang Feng, Lingxi Chen, Liang Zhang
Förderung
- Institute of Digital Medicine
- Tung Foundation Biomedical Sciences Centre
- Research Grants Council of Hong Kong
- Shanghai Action Plan for Science, Technology and Innovation Field of Computational Biology
- Lingang Laboratory
- National Natural Science Foundation of China
Zitieren
APA
Guo, W., Jin, W., Zheng, J., Zhang, J., Feng, X., Chen, L., & Zhang, L. (2026). Biologically informative not available deconvolution excavates hidden features in proteomics datasets. Briefings in Bioinformatics. https://doi.org/10.1093/bib/bbag582
Vancouver
Guo W, Jin W, Zheng J, Zhang J, Feng X, Chen L, et al. Biologically informative not available deconvolution excavates hidden features in proteomics datasets. Briefings in Bioinformatics. 2026. doi:10.1093/bib/bbag582
BibTeX
@article{Guo2026,
author = {Guo, Weiheng and Jin, Wenyi and Zheng, Jieyi and Zhang, Jian and Feng, Xikang and Chen, Lingxi and Zhang, Liang},
title = {Biologically informative not available deconvolution excavates hidden features in proteomics datasets},
journal = {Briefings in Bioinformatics},
year = {2026},
doi = {10.1093/bib/bbag582}
}
EUVIMED