PRISM-Code/pyproject.toml at main · HuangLab-PKU/PRISM-Code · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
[build-system]
requires = ["setuptools>=64", "wheel"]
build-backend = "setuptools.build_meta"

[project]
name = "prism"
version = "0.1.0"
description = "PRISM post-stitching analysis: spot detection / readout, gene calling (GMM, codebook-GMM, PoSTcode), and DAPI-based cell segmentation. Image acquisition (focal stacking, illumination correction, registration, stitching) lives in the companion spatial_img_core package (available on request). Chang et al., Nature Biotechnology 2025, doi:10.1038/s41587-025-02883-7."
readme = "README.md"
requires-python = ">=3.10"
license = { file = "LICENSE" }
authors = [{ name = "HuangLab-PKU" }]
classifiers = [
    "Programming Language :: Python :: 3",
    "Programming Language :: Python :: 3.10",
    "Programming Language :: Python :: 3.11",
    "Operating System :: OS Independent",
    "Topic :: Scientific/Engineering :: Bio-Informatics",
]
# Baseline covers intensity readout, the GMM and codebook-GMM gene-calling
# methods, and DAPI segmentation via the cellpose path. Image-processing
# primitives (focal stacking, illumination correction, registration, stitching)
# now live in the companion spatial_img_core package — install it separately when
# the upstream pipeline is needed. PoSTcode (Bayesian gene calling) and
# StarDist-based segmentation remain optional extras (heavyweight DL frameworks
# with conflicting backends).
dependencies = [
    "numpy",
    "scipy",
    "pandas",
    "scikit-learn",
    "scikit-image",
    "opencv-python",
    "tifffile",
    "Pillow",
    "tqdm",
    "PyYAML",
    "matplotlib",
    "seaborn",
    "cmap",
    "pathos",
    "csbdeep",
]

# Optional extras. PoSTcode (HuangLab fork) is NOT on PyPI — install it as
# editable from the SPRINTseq vendored tree:
#     pip install -e <SPRINTseq>/experiments/src/postcode
# StarDist still ships TensorFlow 2.10 wheels and pins numpy<2 on Windows; if
# you need it, build a separate env (see docs/installation.md).
[project.optional-dependencies]
spotiflow = ["spotiflow"]                          # default deep-learning spot detector (readout.py uses device='cuda')
postcode = ["torch", "pyro-ppl"]
stardist-tf = ["tensorflow==2.10.1", "stardist"]
cellpose = ["cellpose"]
gui = ["PySimpleGUI"]
test = ["pytest"]

[project.urls]
Homepage = "https://github.com/HuangLab-PKU/PRISM-Code"
Repository = "https://github.com/HuangLab-PKU/PRISM-Code"
Publication = "https://doi.org/10.1038/s41587-025-02883-7"

[tool.setuptools.packages.find]
include = ["prism*"]