From 14af63f1cf6780a94b81b689b4ad1c9d0b0fa76c Mon Sep 17 00:00:00 2001
From: Sage Wright <sage.wright@theiagen.com>
Date: Fri, 19 Jul 2024 13:51:42 +0000
Subject: [PATCH] immediately exclude mutations with failed quality in the
 position warnings

---
 Dockerfile             |  4 ++--
 README.md              |  6 +++---
 tbp_parser/LIMS.py     | 10 +++++++++-
 tbp_parser/__init__.py |  2 +-
 4 files changed, 15 insertions(+), 7 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index 7f13a29..2cc1783 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -3,7 +3,7 @@
 # Shelby Bennett, Erin Young, Curtis Kapsak, & Kutluhan Incekara
 
 ARG SAMTOOLS_VER="1.18"
-ARG TBP_PARSER_VER="1.5.10"
+ARG TBP_PARSER_VER="2.0.0"
 
 FROM ubuntu:jammy as builder
 
@@ -42,7 +42,7 @@ ARG TBP_PARSER_VER
 LABEL base.image="ubuntu:jammy"
 LABEL dockerfile.version="1"
 LABEL software="tbp-parser"
-LABEL software.version="1.5.10"
+LABEL software.version="2.0.0"
 LABEL description="tbp-parser and samtools"
 LABEL website="https://github.com/theiagen/tbp-parser"
 LABEL license="https://github.com/theiagen/tbp-parser/blob/main/LICENSE"
diff --git a/README.md b/README.md
index b20e611..b5dc74d 100644
--- a/README.md
+++ b/README.md
@@ -22,16 +22,16 @@ See also [this page](https://theiagen.notion.site/tbp-parser-b02bef0cbc814b12987
 We highly recommend using the following Docker image to run tbp-parser:
 
 ```markdown
-docker pull us-docker.pkg.dev/general-theiagen/theiagen/tbp-parser:1.5.10
+docker pull us-docker.pkg.dev/general-theiagen/theiagen/tbp-parser:2.0.0
 ```
 
 The entrypoint for this Docker image is the tbp-parser help message. To run this container interactively, use the following command:
 
 ```markdown
-docker run -it --entrypoint=/bin/bash us-docker.pkg.dev/general-theiagen/theiagen/tbp-parser:1.5.10
+docker run -it --entrypoint=/bin/bash us-docker.pkg.dev/general-theiagen/theiagen/tbp-parser:2.0.0
 # Once inside the container interactively, you can run the tbp-parser tool
 python3 /tbp-parser/tbp_parser/tbp_parser.py -v
-# v1.5.10
+# v2.0.0
 ```
 
 ### Locally with Python
diff --git a/tbp_parser/LIMS.py b/tbp_parser/LIMS.py
index 22f6b62..558bf77 100644
--- a/tbp_parser/LIMS.py
+++ b/tbp_parser/LIMS.py
@@ -241,6 +241,10 @@ def apply_lims_rules(self, gene_dictionary, DF_LIMS, max_mdl_resistance, antimic
                
               self.logger.debug("LIMS:Since this MDL interpretation changed, we are now potentially recalculating max_mdl_resistance (currently {})".format(max_mdl_resistance[0]))
               
+              if (len(all_responsible_mdl_interpretations) == 0):
+                self.logger.debug("LIMS:There are no more potential MDL interpretations; exiting loop")
+                break
+              
               if (max([globals.RESISTANCE_RANKING[interpretation] for gene_set in all_responsible_mdl_interpretations.values() for interpretation in gene_set]) != globals.RESISTANCE_RANKING[max_mdl_resistance[0]]) and gene in responsible_gene:
                 max_mdl_resistance = [annotation for annotation, rank in globals.RESISTANCE_RANKING.items() if rank == max([globals.RESISTANCE_RANKING[interpretation] for gene_set in all_responsible_mdl_interpretations.values() for interpretation in gene_set])]
                 self.logger.debug("LIMS:The maximum needed to be reevaluated; the potential new max_mdl_resistance is now {}".format(max_mdl_resistance[0]))
@@ -353,7 +357,11 @@ def create_lims_report(self):
     for antimicrobial_code, gene_dictionary in globals.ANTIMICROBIAL_CODE_TO_GENES.items():
       drug_name = globals.ANTIMICROBIAL_CODE_TO_DRUG_NAME[antimicrobial_code]
       # get the MDL interpretations for all genes **FOR THE LIMS REPORT** associated with this drug             
-      potential_mdl_resistances = globals.DF_LABORATORIAN[globals.DF_LABORATORIAN["antimicrobial"] == drug_name].loc[globals.DF_LABORATORIAN["tbprofiler_gene_name"].isin(gene_dictionary.keys())]["mdl_interpretation"]
+      potential_mdl_resistances = globals.DF_LABORATORIAN[globals.DF_LABORATORIAN["antimicrobial"] == drug_name].loc[globals.DF_LABORATORIAN["tbprofiler_gene_name"].isin(gene_dictionary.keys())]
+
+      # remove any interpretations with failed quality warnings
+      potential_mdl_resistances = potential_mdl_resistances.loc[~potential_mdl_resistances["warning"].str.contains("Failed quality in the mutation position")]
+      potential_mdl_resistances = potential_mdl_resistances["mdl_interpretation"].tolist()
       
       # initalize list of genes responsible for the max resistance
       responsible_gene = set()
diff --git a/tbp_parser/__init__.py b/tbp_parser/__init__.py
index 5f5bf9a..e20a8c8 100644
--- a/tbp_parser/__init__.py
+++ b/tbp_parser/__init__.py
@@ -1 +1 @@
-__VERSION__ = "v1.5.10"
\ No newline at end of file
+__VERSION__ = "v2.0.0"
\ No newline at end of file