library(ROMOPAPI)
test_databasesConfig <- HadesExtras_readAndParseYaml(
pathToYalmFile = system.file("testdata", "config", "onlyCounts_databasesConfig.yml", package = "ROMOPAPI"),
pathToFinnGenCountsSqlite = helper_FinnGen_getDatabaseFileCounts()
)
cohortTableHandlerConfig <- test_databasesConfig[[1]]$cohortTableHandler
# Create CDMdbHandler
CDMdbHandler <- HadesExtras_createCDMdbHandlerFromList(cohortTableHandlerConfig, loadConnectionChecksLevel = "basicChecks")
#> Connecting using SQLite driverBackground: Hierarchy can be expored in Athena, or in Atlas, in Vocabulary > Concept > Hierarchy. However, this shows only the imediate descendants (children) and includes all concepts in the vocabulary, when we may want only these present in the database.
For example, what are the descendants of the standard concept SNOMED ‘Asthma’ with concept id 317009?
results_snomed_asthma <- getCodeCounts(CDMdbHandler, conceptId = 317009)
#> getCodeCounts: Getting code counts for conceptId: 317009
#> getConceptsWithCodeCounts: Getting concepts with code counts
mermaid <- createMermaidGraphFromResults(results_snomed_asthma)
DiagrammeR::mermaid(mermaid, width = "100%", height = "100%")Same can be explore in a table format.
createCodeCountsTableFromResults(results_snomed_asthma)If, the ancestor table has been created also for non-standard concepts, this can be used also for non-standard concepts.
For example, what are the descendants of the non-standard concept ICD10 ‘Asthma’ with concept id 45596282?
#results_icd10_asthma <- getCodeCounts(CDMdbHandler, conceptId = 45596282)
# mermaid <- createMermaidGraphFromResults(results_icd10_asthma)
# DiagrammeR::mermaid(mermaid, width = "100%", height = "100%")Background: Incidence rates can be expored in Atlas, in Data Sources. However, this shows only the incidence rates for selected concept only and does not show the contribution of each descendant concept to the incidence rate.
For the Asthma SNOMED concept, example.
createPlotFromResults(results_snomed_asthma, showsMappings = FALSE, width = 800, height = 800)
#> sum of lineData: 481732
#> sum of stackAreasData: 481732
#> time diff: 0
#> Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
#> ℹ Please use `linewidth` instead.
#> ℹ The deprecated feature was likely used in the ROMOPAPI package.
#> Please report the issue at
#> <https://github.com/javier-gracia-tabuenca-tuni/ROMOPAPI/issues>.
#> This warning is displayed once every 8 hours.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.The plor show with a black line the total number of descendant record counts for the Asthma SNOMED concept code.
The stacked areas show the number of record counts for each descendant concept, over time.
Background: Mapped concepts can be expored in
Athena, or in Atlas, in Vocabulary > Concept > Relationships, or
by creating a Concept Set.
However, this shows all the mapped concepts, not only the ones that are
present in the database.
For the Asthma SNOMED concept, example.
We want to know what non-standard concepts map not only to the Asthma SNOMED concept, but also to its descendants.
mermaid <- createMermaidGraphFromResults(results_snomed_asthma, showsMappings = TRUE)
DiagrammeR::mermaid(mermaid, width = "100%", height = "100%")Same can be explore in a table format by opening the arrows.
createCodeCountsTableFromResults(results_snomed_asthma)Background: This cannot be done in Atlas.
For the Asthma SNOMED concept, example.
createPlotFromResults(results_snomed_asthma, showsMappings = TRUE, width = 800, height = 800)
#> sum of lineData: 481732
#> sum of stackAreasData: 481732
#> time diff: 0In this case the stacked areas show the number of record counts for each source concept, over time.
Sometimes the tree of descendant may become to granular.
For example, in the case of drugs, if take ATC level 4 C10AA for statins with concept id 21601855, the tree and plot of descendant may become to detailed, as it takes all the lowest rxnorm packages concepts.
results_atc_c10aa <- getCodeCounts(CDMdbHandler, conceptId = 21601855)
#> getCodeCounts: Getting code counts for conceptId: 21601855
mermaid <- createMermaidGraphFromResults(results_atc_c10aa)
DiagrammeR::mermaid(mermaid, width = "100%", height = "100%")If we are interested only in the inmediate descendants ATC level 5s, we can prune the tree by setting pruneLevels to 2.
results_atc_c10aa_pruned <- pruneLevelsFromResults(results_atc_c10aa, pruneLevels = 2)
#> Warning: There was 1 warning in `dplyr::mutate()`.
#> ℹ In argument: `min_level = dplyr::if_else(...)`.
#> Caused by warning in `dplyr::if_else()`:
#> ! NAs introduced by coercion
mermaid <- createMermaidGraphFromResults(results_atc_c10aa_pruned)
DiagrammeR::mermaid(mermaid, width = "100%", height = "100%")However, this is still including the Ingredient concepts in addition to the ATC level 5s. Which mostly, include the same descendant that the ATC level 5s. This becames a problem if we want to plot the contribution over time, as the stacked areas will include both the descendant of the Ingredient concepts and the ATC level 5s. This makes the stacked areas to be double of the root node descendant record counts (black line).
createPlotFromResults(results_atc_c10aa_pruned, showsMappings = FALSE, width = 800, height = 800)
#> sum of lineData: 6209837
#> sum of stackAreasData: 12419674
#> time diff: 6209837If we are interested only in the ATC level 5s, we can prune the tree by setting pruneClass to “Ingredient”.
results_atc_c10aa_pruned_ingredient <- pruneLevelsFromResults(results_atc_c10aa, pruneLevels = 2, pruneClass = "Ingredient")
#> Warning: There was 1 warning in `dplyr::mutate()`.
#> ℹ In argument: `min_level = dplyr::if_else(...)`.
#> Caused by warning in `dplyr::if_else()`:
#> ! NAs introduced by coercion
mermaid <- createMermaidGraphFromResults(results_atc_c10aa_pruned_ingredient)
DiagrammeR::mermaid(mermaid, width = "100%", height = "100%")The plot will now show only the ATC level 5s and now the stacked areas will be the same as the root node descendant record counts (black line).
createPlotFromResults(results_atc_c10aa_pruned_ingredient, showsMappings = FALSE, width = 800, height = 800)
#> sum of lineData: 6209837
#> sum of stackAreasData: 6209837
#> time diff: 0