New Metrics: Kendal tau and Geodestic distance

CHANGELOG.md

@@ -9,7 +9,7 @@
 * Added `spaTrack` method (PR #4).
 * Added `Spearman's correlation` metric (PR #5).
 * Added `Moran's I` metric (PR #6).
-
+* Added `kendall_tau` and `geodesic distance` metric (PR #8)
 * Added `stlearn`method (PR #9).
 
 ## MAJOR CHANGES

src/metrics/geodesic_distance/config.vsh.yaml

@@ -0,0 +1,64 @@
+__merge__: ../../api/comp_metric.yaml
+
+name: geodesic_distance
+
+info:
+  metrics:
+
+    - name: geodesic_distance
+      label: Geodesic Distance Correlation
+      summary: "Computes the Spearman correlation between pairwise geodesic distances of inferred and ground-truth pseudotime."
+      description: |
+        Calculates the Spearman rank correlation between all pairwise distances of predicted pseudotime and all pairwise distances of ground-truth pseudotime, computed over every unique pair of cells.
+        This measures whether the relative spacing between cells along the trajectory is preserved and captures distortions where a method compresses or stretches segments of the trajectory even if the overall cell ordering is correct.
+      references:
+        doi:
+          - 10.1038/s41592-020-0772-5
+        bibtex: |
+          @article{Virtanen_2020,
+           author  = {Virtanen, Pauli and Gommers, Ralf and Oliphant, Travis E. and Haberland, Matt and Reddy, Tyler and Cournapeau, David and Burovski, Evgeni and Peterson, Pearu and Weckesser, Warren and Bright, Jonathan and {van der Walt}, St{\'e}fan J. and Brett, Matthew and Wilson, Joshua and Jarrod Millman, K. and Mayorov, Nikolay and Nelson, Andrew R. J. and Jones, Eric and Kern, Robert and Larson, Eric and Carey, C. J. and Polat, {\dot{I}}lhan and Feng, Yu and Moore, Eric W. and VanderPlas, Jake and Laxalde, Denis and Perktold, Josef and Cimrman, Robert and Henriksen, Ian and Quintero, E. A. and Harris, Charles R. and Archibald, Anne M. and Ribeiro, Ant{\^o}nio H. and Pedregosa, Fabian and {van Mulbregt}, Paul and {SciPy 1.0 Contributors}},
+           title   = {Author Correction: SciPy 1.0: fundamental algorithms for scientific computing in Python},
+           journal = {Nature Methods},
+           volume  = {17},
+           number  = {3},
+           pages   = {352},
+           year    = {2020},
+           doi     = {10.1038/s41592-020-0772-5}
+           }
+      links:
+        documentation: https://docs.scipy.org/doc/scipy/reference/spatial.distance.html
+        repository: https://github.com/scipy/scipy
+      min: -1
+      max: 1
+      maximize: true
+
+resources:
+  - type: python_script
+    path: script.py
+
+engines:
+  - type: docker
+    image: python:3.11-slim
+    setup:
+      - type: apt
+        packages:
+          - procps              # required by Nextflow
+          - git                 # pip needs it to install openproblems core from git+https
+      - type: python
+        packages:
+          - anndata~=0.10.9
+          - scanpy~=1.10.4
+          - numpy~=2.4.6
+          - pandas~=3.0.5
+          - scipy~=1.17.1
+          - pyyaml~=6.0.3
+          - requests~=2.34.2
+          - jsonschema~=4.26.0
+        github:
+          - "openproblems-bio/core#subdirectory=packages/python/openproblems"
+
+runners:
+  - type: executable
+  - type: nextflow
+    directives:
+      label: [midtime,midmem,midcpu]

src/metrics/geodesic_distance/script.py

@@ -0,0 +1,30 @@
+import anndata as ad
+
+## VIASH START
+# Note: this section is auto-generated by viash at runtime. To edit it, make changes
+# in config.vsh.yaml and then run `viash config inject config.vsh.yaml`.
+par = {
+  'input_solution': 'resources_test/.../solution.h5ad',
+  'input_prediction': 'resources_test/.../prediction.h5ad',
+  'output': 'output.h5ad'
+}
+meta = {
+  'name': 'geodesic_distance'
+}
+## VIASH END
+
+print('Reading input files', flush=True)
+input_solution = ad.read_h5ad(par['input_solution'])
+input_prediction = ad.read_h5ad(par['input_prediction'])
+
+print('Compute metrics', flush=True)
+# metric_ids and metric_values can have length > 1
+# but should be of equal length
+uns_metric_ids = [ 'geodesic_distance' ]
+uns_metric_values = [ 0.5 ]
+
+print("Write output AnnData to file", flush=True)
+output = ad.AnnData(
+
+)
+output.write_h5ad(par['output'], compression='gzip')

src/metrics/kendall_tau/config.vsh.yaml

@@ -0,0 +1,65 @@
+__merge__: ../../api/comp_metric.yaml
+
+name: kendall_tau
+
+info:
+  metrics:
+
+    - name: kendall_tau
+      label: Kendall Tau
+      summary: "Computes the Kendall's τ coefficient between inferred and ground-truth pseudotime."
+      description: |
+        Calculates the Kendall's τ coefficient between predicted and ground-truth pseudotime. 
+        Kendall's τ (tau) is a non-parametric statistic that measures the ordinal association, or rank correlation, between two variables based on how similarly they order a set of observations.
+      references:
+        doi:
+          - 10.1038/s41592-020-0772-5
+        bibtex: |
+          @article{Virtanen_2020,
+           author  = {Virtanen, Pauli and Gommers, Ralf and Oliphant, Travis E. and Haberland, Matt and Reddy, Tyler and Cournapeau, David and Burovski, Evgeni and Peterson, Pearu and Weckesser, Warren and Bright, Jonathan and {van der Walt}, St{\'e}fan J. and Brett, Matthew and Wilson, Joshua and Jarrod Millman, K. and Mayorov, Nikolay and Nelson, Andrew R. J. and Jones, Eric and Kern, Robert and Larson, Eric and Carey, C. J. and Polat, {\dot{I}}lhan and Feng, Yu and Moore, Eric W. and VanderPlas, Jake and Laxalde, Denis and Perktold, Josef and Cimrman, Robert and Henriksen, Ian and Quintero, E. A. and Harris, Charles R. and Archibald, Anne M. and Ribeiro, Ant{\^o}nio H. and Pedregosa, Fabian and {van Mulbregt}, Paul and {SciPy 1.0 Contributors}},
+           title   = {Author Correction: SciPy 1.0: fundamental algorithms for scientific computing in Python},
+           journal = {Nature Methods},
+           volume  = {17},
+           number  = {3},
+           pages   = {352},
+           year    = {2020},
+           doi     = {10.1038/s41592-020-0772-5}
+           }
+      links:
+        documentation: https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.kendalltau.html
+        repository: https://github.com/scipy/scipy
+      min: -1
+      max: 1
+      maximize: true
+
+resources:
+  - type: python_script
+    path: script.py
+
+engines:
+  - type: docker
+    image: python:3.11-slim
+    setup:
+      - type: apt
+        packages:
+          - procps              # required by Nextflow
+          - git                 # pip needs it to install openproblems core from git+https
+      - type: python
+        packages:
+          - anndata~=0.10.9
+          - scanpy~=1.10.4
+          - numpy~=2.4.6
+          - pandas~=3.0.5
+          - scipy~=1.17.1
+          - pyyaml~=6.0.3
+          - requests~=2.34.2
+          - jsonschema~=4.26.0
+        github:
+          - "openproblems-bio/core#subdirectory=packages/python/openproblems"
+
+
+runners:
+  - type: executable
+  - type: nextflow
+    directives:
+      label: [midtime,midmem,midcpu]

src/metrics/kendall_tau/script.py

@@ -0,0 +1,68 @@
+import anndata as ad
+import numpy as np
+import pandas as pd
+from scipy import stats
+
+## VIASH START
+# Note: this section is auto-generated by viash at runtime. To edit it, make changes
+# in config.vsh.yaml and then run `viash config inject config.vsh.yaml`.
+par = {
+  'input_solution': 'resources_test/task_template/cxg_mouse_pancreas_atlas/solution.h5ad',
+  'input_prediction': 'resources_test/task_template/cxg_mouse_pancreas_atlas/prediction.h5ad',
+  'output': 'output.h5ad'
+}
+
+meta = {
+  'name': 'kendall_tau'
+}
+## VIASH END
+
+
+def compute_kendall_tau(true_values, inferred_values):
+    """Calculates Kendall tau rank correlation on finite overlapping values."""
+    mask = np.isfinite(true_values) & np.isfinite(inferred_values)
+
+    if mask.sum() < 2:
+        return 0.0
+    kendall_tau, _ = stats.kendalltau(true_values[mask], inferred_values[mask])
+
+    if np.isnan(kendall_tau):
+        return 0.0
+
+    return float(kendall_tau)
+
+print('Reading input files', flush=True)
+input_solution = ad.read_h5ad(par['input_solution'])
+input_prediction = ad.read_h5ad(par['input_prediction'])
+
+assert (input_prediction.obs_names == input_solution.obs_names).all(), "obs_names not the same in prediction and solution inputs"
+
+# ground truth and predicted pseudotime
+TRUE_COL = "pseudotime_true"
+INFERRED_COL = "pseudotime_inferred"
+true_vals = pd.to_numeric(input_solution.obs[TRUE_COL], errors='coerce').values
+inferred_vals = pd.to_numeric(input_prediction.obs[INFERRED_COL], errors='coerce').values
+
+
+score = compute_kendall_tau(true_vals, inferred_vals)
+
+
+# metric_ids and metric_values can have length > 1
+# but should be of equal length
+uns_metric_ids = [ 'kendall_tau' ]
+uns_metric_values = [ score ]
+
+print("Write output AnnData to file", flush=True)
+output = ad.AnnData(
+    obs=pd.DataFrame(index=pd.Index(np.array([], dtype=str))),
+    var=pd.DataFrame(index=pd.Index(np.array([], dtype=str))),
+    uns={
+        'dataset_id': input_solution.uns.get('dataset_id', 'unknown'),
+        'normalization_id': input_solution.uns.get('normalization_id', 'unknown'),
+        'method_id': input_prediction.uns.get('method_id', 'unknown'),
+        'metric_ids': uns_metric_ids,
+        'metric_values': uns_metric_values,
+    }
+)
+
+output.write_h5ad(par['output'], compression='gzip')