Got working connection between trials and NDA/ANDA start times

.dbeaver/.data-sources.json.bak

@@ -0,0 +1,46 @@
+{
+	"folders": {},
+	"connections": {
+		"postgres-jdbc-186be7b92d8-4d6ef891885b76f5": {
+			"provider": "postgresql",
+			"driver": "postgres-jdbc",
+			"name": "aact_db",
+			"save-password": true,
+			"configuration": {
+				"host": "will-office",
+				"port": "5432",
+				"database": "aact_db",
+				"url": "jdbc:postgresql://will-office:5432/aact_db",
+				"configurationType": "MANUAL",
+				"type": "dev",
+				"auth-model": "native"
+			}
+		},
+		"mariaDB-186be7ba5a8-7a57b538dcd4188f": {
+			"provider": "mysql",
+			"driver": "mariaDB",
+			"name": "rxnorm_current",
+			"save-password": true,
+			"configuration": {
+				"host": "will-office",
+				"port": "3306",
+				"database": "rxnorm_current",
+				"url": "jdbc:mariadb://will-office:3306/rxnorm_current",
+				"configurationType": "MANUAL",
+				"type": "dev",
+				"auth-model": "native"
+			}
+		}
+	},
+	"connection-types": {
+		"dev": {
+			"name": "Development",
+			"color": "255,255,255",
+			"description": "Regular development database",
+			"auto-commit": true,
+			"confirm-execute": false,
+			"confirm-data-change": false,
+			"auto-close-transactions": false
+		}
+	}
+}

.dbeaver/.project-metadata.json.bak

@@ -0,0 +1 @@
+{"resources":{"development_sql/ASSOICATING NCTIDs to NDCs and Marketing dates.sql":{"default-datasource":"postgres-jdbc-186be7b92d8-4d6ef891885b76f5","default-catalog":"aact_db","default-schema":"public"}}}

.project

@@ -1,11 +1,12 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <projectDescription>
-	<name>development_sql</name>
+	<name>ClinicalTrialsDataProcessing</name>
 	<comment></comment>
 	<projects>
 	</projects>
 	<buildSpec>
 	</buildSpec>
 	<natures>
+		<nature>org.jkiss.dbeaver.DBeaverNature</nature>
 	</natures>
 </projectDescription>

development_sql/ASSOICATING NCTIDs to NDCs and Marketing dates.sql

@@ -1,91 +1,79 @@
 
-/*get all the ndc codes associated with an rxcui
+/* OVERVIEW
- * Same as query
- * http://will-office:4000/REST/rxcui/1668240/allhistoricalndcs.json
- * note the different formats of the dates.
  * 
- * Based on http://will-office:4000/RxNav/search?searchBy=RXCUI&searchTerm=1668240
+ * This links trials to the first date each drug (indexed by NDA/ANDA etc) is
- * it appears that this rxcui is a sbd or bpck (branded drug or pack)
+ * put on the market.
  * 
-
+ * It takes 3 views to build up to it.
- * If I grab every brand, then every branded drug or pack associated with that drug and then attach that to the nsde data I would get the marketing dates required.
- * --get brand names
- * trial -> mesh_term -> IN/MIN (rxcui) -> BN (rxcui) 
- * -- associate brand names to marketing dates
- * BN (rxcui) --> SBD/BPCK (RXCUI) --> ndc11 --> nsde 
  * */
 
-/*
+--Match trials to brands and ingredients
- * I do need to figure out a way to change the date types when importing into postgres. In mariadb they ar mmYYYY wheras in the jsonapi they are YYYYmm but I want is YYYY-mm-01
+create or replace view public.match_trials_to_bn_in as
- */
-
----assoicate ingredients, brands, and approved packaging RXCUIs.
-create temp table trial_to_pack_rxcui as
 with trialncts as (
         SELECT DISTINCT nct_id  FROM history.trial_snapshots TS
     )
 SELECT 
     bi.nct_id ,
     bi.downcase_mesh_term, 
-    rp.rxcui  AS drug_rxcui,
+    rr.tty2 ,
-    rn.tty1,
+    rr.rxcui2 as bn_or_in_cui --brand or ingredient
-    rn.rxcui1 as ingredient_rxcui, --ingredients
+    ,count(*)
-    rn.tty2 ,
-    rn.rxcui2 as brand_or_pack_rxcui --brand or pack
 FROM ctgov.browse_interventions bi
-    LEFT OUTER JOIN rxnorm_migrated.rxnorm_props AS rp
+    left outer JOIN rxnorm_migrated.rxnorm_props AS rp
-        on bi.downcase_mesh_term  = rp.propvalue1  --Link drug ingredient
+        on bi.downcase_mesh_term  = rp.propvalue1 --link names to drug cuis ()
-    LEFT OUTER JOIN rxnorm_migrated.rxnorm_relations AS rn
+    left outer join rxnorm_migrated.rxnorm_relations rr 
-        on rp.rxcui = rn.rxcui1 --Grab brand names
+    	on rr.rxcui1 = rp.rxcui 
 WHERE
     bi.nct_id in (
         SELECT nct_id  FROM trialncts
         )
     and
     bi.mesh_type='mesh-list'
-    and
+    and rp.propname = 'Active_ingredient_name'
-    rn.tty1 in ('IN','MIN')
+    and rr.tty2 in ('BN', 'IN', 'MIN')
-    and 
+group by bi.nct_id, bi.downcase_mesh_term , rr.tty2 ,rr.rxcui2 
-    rn.tty2 = 'BN'
+order by bi.nct_id 
 ;
 
 
---link brand names to drug applications (NDA/ANDA etc)
+--match trials to through brands NDC11
-select rr.rxcui1 as BN, rr.rxcui2 as pack, ah.ndc as pack_ndc11
+create or replace view public.match_trial_to_ndc11 as
-from 
+select 
-    rxnorm_migrated.rxnorm_relations rr 
+	mttbi.nct_id,
+	ah.ndc,
+	count(*)
+from public.match_trials_to_bn_in as mttbi
+    left outer join rxnorm_migrated.rxnorm_relations as rr 
+    	on mttbi.bn_or_in_cui = rr.rxcui1
     left outer join rxnorm_migrated."ALLNDC_HISTORY" as ah 
         on rr.rxcui2 = ah.rxcui 
 where 
-    tty1 = 'BN' 
+    rr.tty1 = 'BN' 
     and
-    tty2 in ('SBD', 'BPCK')
+    rr.tty2 in ('SBD', 'BPCK')
     and 
     ah.sab='RXNORM' 
-    
+group by mttbi.nct_id, ah.ndc
+order by mttbi.nct_id, ah.ndc
 ;
 
 
----associate NDAs/ANDAs to marketing start dates
----Get start of coverage periods for NSDE dates grouped by arbitrary grouping.
-SELECT n.application_number_or_citation, count(*), min( marketing_start_date )
-FROM spl.nsde as n
-where product_type = 'HUMAN PRESCRIPTION DRUG'
-group by n.application_number_or_citation ;
-
 
----For a given date, find which NDAs/ANDAs were active were active.
+---associate trials to marketing start dates
-SELECT n.application_number_or_citation, count(*)
+create or replace view public.match_trial_to_marketing_start_date as
-FROM spl.nsde as n
+select 
+	mttn.nct_id,
+	n.application_number_or_citation,
+	min(n.marketing_start_date )
+from match_trial_to_ndc11 mttn 
+	inner join spl.nsde n 
+		on mttn.ndc = n.package_ndc11 
 where 	
-    product_type = 'HUMAN PRESCRIPTION DRUG'
+	n.product_type = 'HUMAN PRESCRIPTION DRUG'
-    and
+	and
-    marketing_start_date < '2010-05-01'
+	n.marketing_category in ('NDA','ANDA','BLA', 'NDA authorized generic', 'NDA AUTHORIZED GENERIC')
-    and
+group by mttn.nct_id,n.application_number_or_citation 
-    marketing_end_date > '2010-05-01'
+order by mttn.nct_id 
-group by n.application_number_or_citation ;
+;
-
-
-
 
 

development_sql/views to history.sql

@@ -1,5 +1,5 @@
 /***************CREATE VIEWS*******************/
-create view if not exists 
+create or replace view 
 	history.match_drugs_to_trials as
 select nct_id,  rxcui, propvalue1
 from