task_batch_integration/src/methods/density_adaptive/config.vsh.yaml at 7f0e5a501f85657124d24d95a20837e7d7bf172f · openproblems-bio/task_batch_integration · 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
65
66
67
68
69
70
71
72
73
74
75
76
# The API specifies which type of component this is.
# It contains specifications for:
#   - The input/output files
#   - Common parameters
#   - A unit test
__merge__: ../../api/comp_method.yaml

name: density_adaptive
label: Density-Adaptive BBSG
summary: "Density-adaptive batch-balanced similarity graph with variance-weighted features"
description: |
  A batch integration method using variance-weighted PCA, Combat residualization,
  and a density-adaptive batch-balanced similarity graph (BBSG) that adjusts
  cross-batch connectivity based on local density patterns.
references:
  bibtex:
    - |
      @article{chung2025station,
        title={The Station: An Open-World Environment for AI-Driven Discovery},
        author={Chung, Stephen and Du, Wenyu},
        journal={arXiv preprint arXiv:2511.06309},
        year={2025}
      }
links:
  documentation: https://github.com/dualverse-ai/station/tree/main/example/research_batch_integration/misc/station_sota
  repository: https://github.com/dualverse-ai/station

info:
  method_types: [embedding, graph]
  preferred_normalization: counts

arguments:
  - name: --n_hvgs
    type: integer
    default: 1500
    description: Number of highly variable genes to use
  - name: --k_total
    type: integer
    default: 48
    description: Total neighbors per cell in BBSG
  - name: --k_density
    type: integer
    default: 20
    description: Neighbor rank for local density proxy
  - name: --delta
    type: double
    default: 0.10
    description: Max modulation of cross-batch fraction (e.g., 0.15 = ±15%)
  - name: --alpha_var_graph
    type: double
    default: 0.6
    description: Variance weighting exponent for graph construction
  - name: --alpha_var_emb
    type: double
    default: 0.5
    description: Variance weighting exponent for embedding

resources:
  - type: python_script
    path: script.py

engines:
  - type: docker
    image: openproblems/base_python:1
    setup:
      - type: python
        pypi:
          - scanpy
          - scikit-learn
          - scikit-misc

runners:
  - type: executable
  - type: nextflow
    directives:
      label: [hightime, highmem, midcpu]