fix: collapse multi-site variant names

workflow/scripts/fill_all_sites.R

@@ -1,6 +1,4 @@
 #!/usr/bin/env Rscript
-
-# Write stdout and stderr to log file
 log <- file(snakemake@log[[1]], open = "wt")
 sink(log, type = "message")
 sink(log, type = "output")
@@ -15,10 +13,18 @@ log_threshold(INFO)
 log_info("Reading variants")
 variants <- read_tsv(snakemake@input[["variants"]])
 
-# Create a mapping of variant names to their genomic position
-variant_coords <- variants %>%
-  select(VARIANT_NAME, CHROM, POS) %>%
-  distinct()
+# Check if each sample, variant and position have 1 frequency
+variants %>%
+  distinct(VARIANT_NAME, CHROM, POS, SAMPLE, ALT_FREQ) %>%
+  group_by(VARIANT_NAME, CHROM, POS, SAMPLE) %>%
+  filter(n() > 1) %>%
+  {
+    if (nrow(.) > 0) {
+      log_warn(
+        "Found {nrow(.)} ambiguous (SAMPLE, VARIANT_NAME, POS) combinations"
+      )
+    }
+  }
 
 log_info("Reading filtered sites")
 sites <- read_tsv(snakemake@input[["sites"]]) %>%
@@ -28,14 +34,9 @@ sites <- read_tsv(snakemake@input[["sites"]]) %>%
 log_info("Processing variants")
 all_variants <- variants %>%
   # Select minimal columns
-  distinct(VARIANT_NAME, CHROM, SAMPLE, ALT_FREQ) %>%
-  # Handle duplicates
-  group_by(SAMPLE, VARIANT_NAME, CHROM) %>%
-  summarise(ALT_FREQ = sum(ALT_FREQ, na.rm = TRUE), .groups = "drop") %>%
+  distinct(VARIANT_NAME, CHROM, POS, SAMPLE, ALT_FREQ) %>%
   # Complete with NA
-  complete(SAMPLE, VARIANT_NAME, CHROM) %>%
-  # Assign genomic positions for all combinations
-  left_join(variant_coords, by = c("CHROM", "VARIANT_NAME")) %>%
+  complete(SAMPLE, nesting(VARIANT_NAME, CHROM, POS)) %>%
   # Merge filtered sites
   # TODO: consider region/chrom
   left_join(sites, by = c("SAMPLE", "POS")) %>%

workflow/scripts/report/pairwise_trajectory_correlation_data.R

@@ -1,6 +1,4 @@
 #!/usr/bin/env Rscript
-
-# Write stdout and stderr to log file
 log <- file(snakemake@log[[1]], open = "wt")
 sink(log, type = "message")
 sink(log, type = "output")
@@ -17,14 +15,22 @@ log_threshold(INFO)
 log_info("Reading variants")
 variants <- read_tsv(snakemake@input[["variants"]])
 
-# Obtain sample names ordered by CollectionDate
+log_info("Sorting dates")
 date_order <- read_csv(snakemake@input[["metadata"]]) %>%
   arrange(CollectionDate) %>%
   pull(ID) %>%
   unique()
 
 log_info("Formatting variants")
 all_variants_wider <- variants %>%
+  # Collapse positions (treat as uncertain if filter is inconsistent)
+  group_by(SAMPLE, VARIANT_NAME) %>%
+  summarise(
+    ALT_FREQ = ifelse(n_distinct(ALT_FREQ, na.rm = TRUE) > 1, NA, first(ALT_FREQ)),
+    FILTER_PASS = ifelse(n_distinct(FILTER_PASS) > 1, NA, first(FILTER_PASS)),
+    .groups = "drop"
+  ) %>%
+  mutate(ALT_FREQ = if_else(is.na(FILTER_PASS), NA, ALT_FREQ)) %>%
   distinct(SAMPLE, VARIANT_NAME, ALT_FREQ) %>%
   pivot_wider(
     names_from = VARIANT_NAME,