Metadata-Version: 2.1
Name: bio-past
Version: 1.0.0
Summary: PAST: latent feature extraction with a Prior-based self-Attention framework for Spatial Transcriptomics
Home-page: https://github.com/lizhen18THU/PAST
Author: Zhen Li
Author-email: lizhen18THU@163.com
License: MIT License
Keywords: pip,past
Platform: UNKNOWN
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3.8
Classifier: Operating System :: MacOS :: MacOS X
Classifier: Operating System :: Microsoft :: Windows
Classifier: Operating System :: POSIX :: Linux
Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
Requires-Python: >3.6.0
License-File: LICENSE
Requires-Dist: numba (>=0.55.1)
Requires-Dist: numpy (>=1.21.6)
Requires-Dist: pandas (>=1.3.4)
Requires-Dist: scipy (>=1.8.1)
Requires-Dist: scikit-learn (>=1.1.1)
Requires-Dist: scanpy (>=1.9.1)
Requires-Dist: torch (>=1.9.1)
Requires-Dist: rpy2 (>=3.4.5)

PAST: latent feature extraction with a Prior-based self-Attention framework for Spatial Transcriptomics. Recent development of spatial transcriptomics (ST), which can not only obtain comprehensive gene expression profiles but also preserve spatial information, provides a new dimension to genomics research. As a prerequisite and basis for various downstream missions, latent feature extraction is of great significance for the analysis of spatial transcriptomics. However, few methods consider facilitating data analysis via integrating rich prior information from existing data, and the modality fusion of spatial information and gene expression also remains challenging. Here we propose PAST, a representation learning framework for spatial transcriptomics which takes advantage of prior information with Bayesian neural network and integrates spatial information and gene profile with self-attention mechanism

