Refine handling of positions with deletions and add respective options

CMakeLists.txt

@@ -168,7 +168,7 @@ include( "${CMAKE_CURRENT_LIST_DIR}/tools/cmake/DownloadDependency.cmake" )
 # These are replaced by tools/cmake/update_dependencies.sh to the hashes that are currently checked out.
 # Thus, do not replace the hashes manually!
 SET( CLI11_COMMIT_HASH "5cb3efabce007c3a0230e4cc2e27da491c646b6c" ) #CLI11_COMMIT_HASH#
-SET( genesis_COMMIT_HASH "9a25dfeffcb4fea481c5437725904ed4514da559" ) #genesis_COMMIT_HASH#
+SET( genesis_COMMIT_HASH "1b12550964abb4c2bc421cd9547fe149b563830e" ) #genesis_COMMIT_HASH#
 
 # Call the github download function, which takes four arguments:
 #  - LIBPATH    : Path to the libracy dir where dependencies are stored.

src/commands/analyze/fst.cpp

@@ -123,7 +123,7 @@ void setup_fst( CLI::App& app )
         "--method",
         options->method.value,
         "FST method to use for the computation.\n(1) The unbiased pool-sequencing statistic "
-        "(in two variants, following the definition of Nei, and the definition of Hudson et al),"
+        "(in two variants, following the definition of Nei, and the definition of Hudson),"
         "\n(2) the statistic by Kofler et al of PoPoolation2, or \n(3) the asymptotically unbiased "
         "estimator of Karlsson et al (which is also implemented in PoPoolation2). "
         "\nAll except for the Karlsson method also require `--pool-sizes` to be provided."
@@ -836,9 +836,9 @@ void run_fst( FstOptions const& options )
     );
     genesis::population::VariantFilter total_filter;
     total_filter.only_snps = true;
-    if( state.method == FstMethod::kKarlsson ) {
-        total_filter.only_biallelic_snps = true;
-    }
+    // if( state.method == FstMethod::kKarlsson ) {
+    //     total_filter.only_biallelic_snps = true;
+    // }
     options.variant_input.add_combined_filter_and_transforms(
         options.filter_numerical.make_total_filter( total_filter )
     );

src/options/variant_filter_numerical.cpp

@@ -54,9 +54,6 @@ void VariantFilterNumericalOptions::add_sample_filter_opts_to_app(
     add_sample_count_filter_opts_to_app(    sub, true, true, group );
     add_sample_coverage_filter_opts_to_app( sub, true, true, group );
     add_sample_snp_filter_opts_to_app(      sub, true, true, group );
-
-    // TODO include?!
-    // CliOption<bool> sample_tolerate_deletions  = false;
 }
 
 
@@ -73,7 +70,9 @@ void VariantFilterNumericalOptions::add_sample_count_filter_opts_to_app(
             sample_min_count.value,
             "Minimum base count for a nucleotide (in `ACGT`) to be considered as an allele. "
             "Counts below that are set to zero, and hence ignored as an allele/variant. "
-            "For example, singleton read sequencing errors can be filtered out this way."
+            "For example, singleton read sequencing errors can be filtered out this way. "
+            "This min count is also used for filtering base counts with "
+            "too many deletions, unless `--filter-sample-tolerate-deletions` is set."
         );
         sample_min_count.option->group( group );
     }
@@ -89,6 +88,15 @@ void VariantFilterNumericalOptions::add_sample_count_filter_opts_to_app(
         );
         sample_max_count.option->group( group );
     }
+
+    // Add tolerate deletions flag
+    sample_tolerate_deletions.option = sub->add_flag(
+        "--filter-sample-tolerate-deletions",
+        sample_tolerate_deletions.value,
+        "By default, we filter out any samples that have too many deletions (based on "
+        "`--filter-sample-min-count`). With this flag however, such samples are retained."
+    );
+    sample_tolerate_deletions.option->group( group );
 }
 
 void VariantFilterNumericalOptions::add_sample_coverage_filter_opts_to_app(
@@ -251,33 +259,43 @@ void VariantFilterNumericalOptions::add_total_snp_filter_opts_to_app(
 
 void VariantFilterNumericalOptions::add_total_snp_count_opts_to_app(
     CLI::App* sub,
-    bool add_total_min_count_for_snp,
-    bool add_total_max_count_for_snp,
+    bool add_total_min_count,
+    bool add_total_max_count,
     std::string const& group
 ) {
     // Add min count for snps filter
-    if( add_total_min_count_for_snp ) {
-        total_min_count_for_snp.option = sub->add_option(
-            "--filter-total-snp-min-count",
-            total_min_count_for_snp.value,
+    if( add_total_min_count ) {
+        total_min_count.option = sub->add_option(
+            "--filter-total-min-count",
+            total_min_count.value,
             "When filtering for positions that are SNPs, use this minimum count (summed across all "
             "samples) to identify what is considered a SNP. Positions where the counts are below "
-            "this are filtered out."
+            "this are filtered out. This min count is also used for filtering positions with "
+            "too many deletions, unless `--filter-total-tolerate-deletions` is set."
         );
-        total_min_count_for_snp.option->group( group );
+        total_min_count.option->group( group );
     }
 
     // Add max count for snps filter
-    if( add_total_max_count_for_snp ) {
-        total_max_count_for_snp.option = sub->add_option(
-            "--filter-total-snp-max-count",
-            total_max_count_for_snp.value,
+    if( add_total_max_count ) {
+        total_max_count.option = sub->add_option(
+            "--filter-total-max-count",
+            total_max_count.value,
             "When filtering for positions that are SNPs, use this maximum count (summed across all "
             "samples) to identify what is considered a SNP. Positions where the counts are above "
             "this are filtered out; probably not relevant in practice, but offered for completeness."
         );
-        total_max_count_for_snp.option->group( group );
+        total_max_count.option->group( group );
     }
+
+    // Add tolerate deletions flag
+    total_tolerate_deletions.option = sub->add_flag(
+        "--filter-total-tolerate-deletions",
+        total_tolerate_deletions.value,
+        "By default, we filter out any positions that have too many deletions (based on "
+        "`--filter-total-min-count`). With this flag however, such positions are retained."
+    );
+    total_tolerate_deletions.option->group( group );
 }
 
 void VariantFilterNumericalOptions::add_total_freq_filter_opts_to_app(
@@ -405,18 +423,22 @@ VariantFilterNumericalOptions::get_total_filter(
         filter.only_biallelic_snps = total_only_biallelic_snps.value;
         any_provided = true;
     }
-    if( total_min_count_for_snp.option && *total_min_count_for_snp.option ) {
-        filter.min_count_for_snp = total_min_count_for_snp.value;
+    if( total_min_count.option && *total_min_count.option ) {
+        filter.min_count = total_min_count.value;
         any_provided = true;
     }
-    if( total_max_count_for_snp.option && *total_max_count_for_snp.option ) {
-        filter.max_count_for_snp = total_max_count_for_snp.value;
+    if( total_max_count.option && *total_max_count.option ) {
+        filter.max_count = total_max_count.value;
         any_provided = true;
     }
     if( total_min_frequency.option && *total_min_frequency.option ) {
         filter.min_frequency = total_min_frequency.value;
         any_provided = true;
     }
+    if( total_tolerate_deletions.option && *total_tolerate_deletions.option ) {
+        filter.tolerate_deletions = total_tolerate_deletions.value;
+        any_provided = true;
+    }
 
     return std::make_pair( filter, any_provided );
 }

src/options/variant_filter_numerical.hpp

@@ -124,8 +124,8 @@ class VariantFilterNumericalOptions
 
     void add_total_snp_count_opts_to_app(
         CLI::App* sub,
-        bool add_total_min_count_for_snp = true,
-        bool add_total_max_count_for_snp = true,
+        bool add_total_min_count = true,
+        bool add_total_max_count = true,
         std::string const& group = "Numerical Filters"
     );
 
@@ -220,9 +220,10 @@ class VariantFilterNumericalOptions
     CliOption<size_t> total_max_coverage         = 0;
     CliOption<bool>   total_only_snps            = false;
     CliOption<bool>   total_only_biallelic_snps  = false;
-    CliOption<size_t> total_min_count_for_snp    = 0;
-    CliOption<size_t> total_max_count_for_snp    = 0;
+    CliOption<size_t> total_min_count            = 0;
+    CliOption<size_t> total_max_count            = 0;
     CliOption<double> total_min_frequency        = 0.0;
+    CliOption<bool>   total_tolerate_deletions   = false;
 
 private: