Merge pull request #3 from JensUweUlrich/abundance_threshold

reporting with different abundance threshold successfully tested
JensUweUlrich · Mar 22, 2024 · 408f558 · 408f558
2 parents 2a789a5 + e0773a1
commit 408f558
Show file tree

Hide file tree

Showing 5 changed files with 12 additions and 7 deletions.
diff --git a/src/main/main.cpp b/src/main/main.cpp
@@ -53,7 +53,7 @@ int main(int argc, char const **argv)
 
 	seqan3::argument_parser top_level_parser{"taxor", argc, argv, seqan3::update_notifications::off,
                                              {"build", "search", "profile"}};
-    top_level_parser.info.version = "0.1.2";
+    top_level_parser.info.version = "0.1.3";
 
     try
     {

diff --git a/src/main/taxor_build.cpp b/src/main/taxor_build.cpp
@@ -48,7 +48,7 @@ using sequence_file_t = seqan3::sequence_file_input<hixf::dna4_traits, seqan3::f
 
 void set_up_subparser_layout(seqan3::argument_parser & parser, taxor::build::configuration & config)
 {
-    parser.info.version = "0.1.0";
+    parser.info.version = "0.1.3";
     parser.info.author = "Jens-Uwe Ulrich";
     parser.info.email = "jens-uwe.ulrich@hpi.de";
     parser.info.short_description = "Creates an HIXF index of a given set of fasta files";

diff --git a/src/main/taxor_profile.cpp b/src/main/taxor_profile.cpp
@@ -15,7 +15,7 @@ namespace taxor::profile
 
 void set_up_subparser_layout(seqan3::argument_parser & parser, taxor::profile::configuration & config)
 {
-    parser.info.version = "0.1.0";
+    parser.info.version = "0.1.3";
     parser.info.author = "Jens-Uwe Ulrich";
     parser.info.email = "jens-uwe.ulrich@hpi.de";
     parser.info.short_description = "Taxonomic profiling of a sample by giving read matching results of Taxor search";
@@ -47,6 +47,11 @@ void set_up_subparser_layout(seqan3::argument_parser & parser, taxor::profile::c
                       "Identifier of the analyzed sample",
                       seqan3::option_spec::required);
 
+    parser.add_option(config.threshold, '\0', "min-abundance",
+                      "Minimum abundance to report (default: 0.001)",
+                      seqan3::option_spec::standard,
+                      seqan3::arithmetic_range_validator{static_cast<double>(0.0), static_cast<double>(1.0)});
+
     parser.add_option(config.threads,
                       '\0', "threads",
                       "The number of threads to use.",
@@ -736,7 +741,7 @@ void tax_profile(taxor::profile::configuration& config)
 
     for (auto & t: tax_abundances)
     {
-        if (t.second < 0.001)
+        if (t.second < config.threshold)
             t.second = 0.0;
     }
     std::map<std::string, taxonomy::Profile_Output> rank_profiles = calculate_higher_rank_abundances(tax_abundances,taxpath);
@@ -752,7 +757,7 @@ void tax_profile(taxor::profile::configuration& config)
 
     for (auto & t: tax_abundances)
     {
-        if (t.second < 0.001)
+        if (t.second < config.threshold)
             t.second = 0.0;
     }
 

diff --git a/src/main/taxor_profile_configuration.hpp b/src/main/taxor_profile_configuration.hpp
@@ -12,7 +12,7 @@ struct configuration
     std::string report_file{};
     std::string sequence_abundance_file{};
     std::string sample_id{};
-    //double threshold{-1.0};
+    double threshold{0.001};
     //double error_rate{0.04};
     uint8_t threads{1u};
     bool output_verbose_statistics{false};

diff --git a/src/main/taxor_search.cpp b/src/main/taxor_search.cpp
@@ -31,7 +31,7 @@ namespace taxor::search
 
 void set_up_subparser_layout(seqan3::argument_parser & parser, taxor::search::configuration & config)
 {
-    parser.info.version = "0.1.0";
+    parser.info.version = "0.1.3";
     parser.info.author = "Jens-Uwe Ulrich";
     parser.info.email = "jens-uwe.ulrich@hpi.de";
     parser.info.short_description = "Queries a file of DNA sequences against an HIXF index";