marker_freq_diff allows users to easily calculate the difference in marker abundance (frequency) between different tissue compartments. For us this is typically looking at tumor vs stroma marker abundance.
marker_freq_diff(
mif,
classifier,
ref_level,
diff_level,
mnames,
overwrite = FALSE
)object of class `mif` created with [spatialTIME::create_mif()]
character that specified which column in the spatial data indicates the different tissue compartments profiled
character found in the `classifier` column of spatial data indicating the first value when calculating frequency difference
character found in the `classifier` column of spatial data indicating the second value when calculating frequency difference
character vector of marker names that are found in the spatial data for which to calculate abundance differences
boolean/logical for whether to overwrite previously calculated marker frequency differences
an object of class `mif` with the `derived$frequency_difference` slot filled
To calculate a p-value associated with the difference in proportion of positive markers identified in different classifier compartments, we implemented the use of the Fisher's Exact test.
mif <- create_mif(clinical_data = example_clinical %>%
dplyr::mutate(deidentified_id = as.character(deidentified_id)),
sample_data = example_summary %>%
dplyr::mutate(deidentified_id = as.character(deidentified_id)),
spatial_list = example_spatial,
patient_id = "deidentified_id",
sample_id = "deidentified_sample")
mif = marker_freq_diff(mif = mif,
classifier = "Classifier.Label",
ref_level = "Tumor",
diff_level = "Stroma",
mnames = c("CD3..FOXP3.", "CD3..CD8.", "CD3..PD1.", "CD3..PD.L1.",
"CD3..Opal.570..Positive", "PD1..Opal.650..Positive"),
overwrite = TRUE)