PHANES: an integrated resource of protein features for the human proteome
- 1Italian Institute of Technology
Abstract
Protein analysis and machine-learning workflows increasingly rely on large-scale numerical representations of proteins. While protein language models provide powerful embeddings, their limited interpretability and high computational cost hinder their use in many biological applications. We present PHANES, a human proteome database and web server providing a comprehensive, biologically interpretable set of features for 20335 proteins. PHANES integrates sequence-derived physicochemical descriptors, structure-based and graph-theoretical features computed from AlphaFold predictions, thermostability indices, and tissue expression specificity. In addition to precomputed features for the human proteome, an integrated web server enables on-demand feature calculation for user-provided protein structures. PHANES supports data export, visualization and comparative analysis at the level of individual proteins and protein groups, bridging the gap between complex numerical representations and biological interpretation while providing an interpretable foundation for exploratory analyses and downstream machine-learning applications. PHANES is freely available at https://tools.tartaglialab.com/phanes.
Related
- memVar: an integrated resource for interpreting human membrane protein variation in molecular and disease contexts
- Retrieval of binding sites across the AlphaFold human proteome using protein language model representations
- Proteome-wide quantification of protein turnover in frog and fly embryos reveals divergent strategies of maternal inheritance
- Retracing the Process of Translation: Proteome-wide mapping of stable transcriptomic predictors of protein abundance in cancer cell lines
- CHORUS: A proteome-scale atlas of membrane interaction interfaces constructed using cross-protein language model embeddings from sequences alone
BibTeX
@misc{vrachnos2026phanes,
title = {{PHANES: an integrated resource of protein features for the human proteome}},
author = {D.-M. Vrachnos and A. Armaos and L. Broglia and M. Monti and J. Fiorentino and G. G. Tartaglia},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.14.756639},
url = {https://www.biorxiv.org/content/10.64898/2026.10.14.756639v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - PHANES: an integrated resource of protein features for the human proteome
AU - D.-M. Vrachnos
AU - A. Armaos
AU - L. Broglia
AU - M. Monti
AU - J. Fiorentino
AU - G. G. Tartaglia
PY - 2026
DA - 2026/10/14
PB - bioRxiv
DO - 10.64898/2026.10.14.756639
UR - https://www.biorxiv.org/content/10.64898/2026.10.14.756639v1
AB - Protein analysis and machine-learning workflows increasingly rely on large-scale numerical representations of proteins. While protein language models provide powerful embeddings, their limited interpretability and high computational cost hinder their use in many biological applications. We present PHANES, a human proteome database and web server providing a comprehensive, biologically interpretable set of features for 20335 proteins. PHANES integrates sequence-derived physicochemical descriptors, structure-based and graph-theoretical features computed from AlphaFold predictions, thermostability indices, and tissue expression specificity. In addition to precomputed features for the human proteome, an integrated web server enables on-demand feature calculation for user-provided protein structures. PHANES supports data export, visualization and comparative analysis at the level of individual proteins and protein groups, bridging the gap between complex numerical representations and biological interpretation while providing an interpretable foundation for exploratory analyses and downstream machine-learning applications. PHANES is freely available at https://tools.tartaglialab.com/phanes.
ER -