From b9497e5cdeef7072364964434fce0fdf36dfe8aa Mon Sep 17 00:00:00 2001
From: Diane Trout <diane@caltech.edu>
Date: Thu, 24 Sep 2015 11:49:18 -0700
Subject: [PATCH] Return read_length for file properties

---
 htsworkflow/submission/aws_submission.py  | 2 ++
 htsworkflow/templates/aws_metadata.sparql | 3 ++-
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/htsworkflow/submission/aws_submission.py b/htsworkflow/submission/aws_submission.py
index f4cd5c0..d6f3cdb 100644
--- a/htsworkflow/submission/aws_submission.py
+++ b/htsworkflow/submission/aws_submission.py
@@ -106,6 +106,8 @@ class AWSSubmission(Submission):
         for row in results:
             if 'platform' in row:
                 row['platform'] = platform_alias[row['platform']]
+            if 'read_length' in row:
+                row['read_length'] = int(row['read_length'])
             flowcell_details = {}
             for term in ['machine', 'flowcell', 'lane', 'barcode']:
                 if term in row:
diff --git a/htsworkflow/templates/aws_metadata.sparql b/htsworkflow/templates/aws_metadata.sparql
index fd99df8..dc2ae32 100644
--- a/htsworkflow/templates/aws_metadata.sparql
+++ b/htsworkflow/templates/aws_metadata.sparql
@@ -6,7 +6,7 @@ PREFIX ncbiTaxon: <http://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?id=>
 PREFIX geoSoft: <http://www.ncbi.nlm.nih.gov/geo/info/soft2.html#>
 PREFIX cells: <http://encodewiki.ucsc.edu/EncodeDCC/index.php/Cell_lines#>
 
-select distinct ?dataset ?replicate ?award ?lab ?lane ?barcode ?submitted_file_name ?file_format ?file_size ?output_type ?md5sum ?flowcell ?machine ?platform
+select distinct ?dataset ?replicate ?award ?lab ?lane ?read_length ?barcode ?submitted_file_name ?file_format ?file_size ?output_type ?md5sum ?flowcell ?machine ?platform
 WHERE {
   <{{submission}}> a submissionOntology:submission ;
                    encode3:dataset ?dataset ;
@@ -23,6 +23,7 @@ WHERE {
         htswlib:library ?library ;
         htswlib:flowcell ?flowcell_url ;
         htswlib:lane_number ?lane ;
+        htswlib:read_length ?read_length ;
         a ?fileClass .
 
   ?flowcell_url htswlib:sequenced_by ?machine ;
-- 
2.30.2