Quick Demo: Reference-based simulation for CODEX spatial proteomics data

import simspace as ss
import pandas as pd

# Step 1: Load the reference dataset. We here provide a sample dataset from Xenium human breast tumor dataset.
ref_meta = pd.read_csv('../data/CODEX_metadata.csv', index_col=0)
ref_omics = pd.read_csv('../data/CODEX_count.csv', index_col=0)

ref_meta.head(3)

	Unnamed: 0.2	Unnamed: 0.1	Unnamed: 0	CellID	centroid_x	centroid_y	Area	perimeter	axis_ratio	tile_h	...	IFNG	Ki67	CD44	PD1	TIM3	necrosis_2clusters	necrosis_3clusters	peri_intra_tumoral	X_centroid	Y_centroid
Unnamed: 0.3
876849	876849	876849	3790987	1	39.842857	2.228571	70.0	32.863961	2.111366	37888	...	N	N	N	N	N	0	0	peri	37927.842857	26882.228571
876850	876850	876850	3790988	2	81.000000	1.500000	28.0	18.000000	1.788854	37888	...	N	N	N	N	N	0	0	peri	37969.000000	26881.500000
876851	876851	876851	3790989	3	94.694118	3.058824	85.0	34.485281	1.467060	37888	...	N	N	N	N	N	0	0	peri	37982.694118	26883.058824

3 rows × 26 columns

ref_omics.head(3)

	DAPI	MPO	Ecadherin	PDL1	CD163	PD1	CD47	GAL3	PARP1	LAG3	...	MLH1	Ki67	CD20	ARID1A	IFNG	CD31	PMS	CD44	PanCytokeratin	CD3e
CellID
1	21.585714	0.028571	0.1	0.928571	0.014286	1.214286	1.128571	32.971428	5.657143	1.900000	...	24.314285	0.900000	5.000000	7.657143	0.900000	0.400000	1.814286	2.828571	1.685714	20.142857
2	20.071428	0.000000	0.0	0.178571	0.428571	0.392857	0.607143	14.000000	0.071429	1.428571	...	5.678571	0.857143	1.928571	0.678571	0.678571	0.142857	0.142857	2.821429	1.928571	1.500000
3	17.858824	0.000000	0.0	0.047059	0.000000	0.235294	0.058824	7.482353	0.564706	1.364706	...	6.764706	0.329412	1.835294	1.141176	0.458824	0.082353	0.200000	0.741176	1.635294	0.717647

3 rows × 31 columns

# Step 2: Fit SimSpace spatial parameters from the reference.
# This step uses the reference metadata and omics data to fit the spatial parameters for the simulation.
# One can do this by using the `ss.fit()` method and some preprocessing, which is shown in details in spaital_fitting.ipynb.
# Here we provide a pre-fitted parameters for the CODEX human endometrial cancer dataset.
# The full dataset will be available upon request.
params = ss.util.load_params('../data/CODEX_fitted_params.json')

# Step 3: Simulate the spatial omics data using the fitted parameters.
# This step uses the fitted parameters to simulate the spatial omics data, just as the reference-free simulation.
sim = ss.util.sim_from_params(
    parameters=params,
    shape=(50, 50),
    custom_neighbor=ss.spatial.generate_offsets(3, method='manhattan'),
    seed=0,
)
# This results is the simulated spatial omics data shown in Fig. 6, with x and y axis switched.
sim.plot(figsize=(5, 5), dpi=100, size=14)

# Step 4 (Optional): Simulate omics data based on the spatial simulation.
# The omics profile can also be fitted from the reference dataset if the reference is available.
# Here we use the `fit_scdesign` method to fit the spatial simulation with the reference dataset,
# which requires the R package 'scDesign3' to be installed. One can find more details in README about the installation.
sim.fit_scdesign(
    '../data/CODEX_count.csv',
    '../data/CODEX_metadata.csv',
    'phenotype',      # Column in metadata that contains the cell type information
    'X_centroid',   # Column in metadata that contains the x coordinate of the cell centroid
    'Y_centroid',   # Column in metadata that contains the y coordinate of the cell centroid
    seed=0,
)

Temporary directory created.
scDesgin fit complete.

# Example of plotting a specific gene, e.g., 'PD1'.
sim.plot('PD1', figsize=(5,5), dpi=100, title='PD1 Expression')

sim.plot('fitted_celltype', figsize=(6,6), dpi=100)